diff --git "a/checkpoint-42462/trainer_state.json" "b/checkpoint-42462/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-42462/trainer_state.json"
@@ -0,0 +1,29756 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 42462,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00023550468654326222,
+      "grad_norm": 3.2309679985046387,
+      "learning_rate": 0.00019995760915642223,
+      "loss": 3.1368,
+      "step": 10
+    },
+    {
+      "epoch": 0.00047100937308652445,
+      "grad_norm": 2.1174814701080322,
+      "learning_rate": 0.00019991050821911358,
+      "loss": 2.6966,
+      "step": 20
+    },
+    {
+      "epoch": 0.0007065140596297867,
+      "grad_norm": 2.325835704803467,
+      "learning_rate": 0.00019986340728180493,
+      "loss": 2.4996,
+      "step": 30
+    },
+    {
+      "epoch": 0.0009420187461730489,
+      "grad_norm": 1.9865907430648804,
+      "learning_rate": 0.00019981630634449625,
+      "loss": 2.7315,
+      "step": 40
+    },
+    {
+      "epoch": 0.0011775234327163111,
+      "grad_norm": 1.900598406791687,
+      "learning_rate": 0.00019976920540718763,
+      "loss": 2.6127,
+      "step": 50
+    },
+    {
+      "epoch": 0.0014130281192595733,
+      "grad_norm": 1.5695827007293701,
+      "learning_rate": 0.00019972210446987895,
+      "loss": 2.5491,
+      "step": 60
+    },
+    {
+      "epoch": 0.0016485328058028356,
+      "grad_norm": 2.3850433826446533,
+      "learning_rate": 0.0001996750035325703,
+      "loss": 2.5412,
+      "step": 70
+    },
+    {
+      "epoch": 0.0018840374923460978,
+      "grad_norm": 2.0210142135620117,
+      "learning_rate": 0.00019962790259526165,
+      "loss": 2.485,
+      "step": 80
+    },
+    {
+      "epoch": 0.00211954217888936,
+      "grad_norm": 1.617475152015686,
+      "learning_rate": 0.000199580801657953,
+      "loss": 2.5976,
+      "step": 90
+    },
+    {
+      "epoch": 0.0023550468654326222,
+      "grad_norm": 1.6759183406829834,
+      "learning_rate": 0.00019953370072064435,
+      "loss": 2.3563,
+      "step": 100
+    },
+    {
+      "epoch": 0.0025905515519758842,
+      "grad_norm": 1.495890498161316,
+      "learning_rate": 0.0001994865997833357,
+      "loss": 2.6193,
+      "step": 110
+    },
+    {
+      "epoch": 0.0028260562385191467,
+      "grad_norm": 2.131957769393921,
+      "learning_rate": 0.00019943949884602705,
+      "loss": 2.5216,
+      "step": 120
+    },
+    {
+      "epoch": 0.0030615609250624087,
+      "grad_norm": 1.818166732788086,
+      "learning_rate": 0.0001993923979087184,
+      "loss": 2.4632,
+      "step": 130
+    },
+    {
+      "epoch": 0.003297065611605671,
+      "grad_norm": 1.854960560798645,
+      "learning_rate": 0.00019934529697140973,
+      "loss": 2.4897,
+      "step": 140
+    },
+    {
+      "epoch": 0.003532570298148933,
+      "grad_norm": 1.624773383140564,
+      "learning_rate": 0.0001992981960341011,
+      "loss": 2.614,
+      "step": 150
+    },
+    {
+      "epoch": 0.0037680749846921956,
+      "grad_norm": 1.9114423990249634,
+      "learning_rate": 0.00019925109509679243,
+      "loss": 2.1351,
+      "step": 160
+    },
+    {
+      "epoch": 0.004003579671235458,
+      "grad_norm": 1.608393669128418,
+      "learning_rate": 0.00019920399415948378,
+      "loss": 2.5292,
+      "step": 170
+    },
+    {
+      "epoch": 0.00423908435777872,
+      "grad_norm": 1.6805859804153442,
+      "learning_rate": 0.00019915689322217513,
+      "loss": 2.5797,
+      "step": 180
+    },
+    {
+      "epoch": 0.004474589044321982,
+      "grad_norm": 1.54912269115448,
+      "learning_rate": 0.00019910979228486648,
+      "loss": 2.6075,
+      "step": 190
+    },
+    {
+      "epoch": 0.0047100937308652445,
+      "grad_norm": 1.4087010622024536,
+      "learning_rate": 0.00019906269134755783,
+      "loss": 2.3585,
+      "step": 200
+    },
+    {
+      "epoch": 0.004945598417408506,
+      "grad_norm": 1.7299461364746094,
+      "learning_rate": 0.00019901559041024918,
+      "loss": 2.3914,
+      "step": 210
+    },
+    {
+      "epoch": 0.0051811031039517685,
+      "grad_norm": 1.4757139682769775,
+      "learning_rate": 0.0001989684894729405,
+      "loss": 2.3994,
+      "step": 220
+    },
+    {
+      "epoch": 0.005416607790495031,
+      "grad_norm": 1.5432946681976318,
+      "learning_rate": 0.00019892138853563188,
+      "loss": 2.5407,
+      "step": 230
+    },
+    {
+      "epoch": 0.005652112477038293,
+      "grad_norm": 1.5481261014938354,
+      "learning_rate": 0.0001988742875983232,
+      "loss": 2.4017,
+      "step": 240
+    },
+    {
+      "epoch": 0.005887617163581555,
+      "grad_norm": 1.5139665603637695,
+      "learning_rate": 0.00019882718666101455,
+      "loss": 2.2548,
+      "step": 250
+    },
+    {
+      "epoch": 0.006123121850124817,
+      "grad_norm": 1.4921780824661255,
+      "learning_rate": 0.00019878008572370593,
+      "loss": 2.293,
+      "step": 260
+    },
+    {
+      "epoch": 0.00635862653666808,
+      "grad_norm": 1.9517822265625,
+      "learning_rate": 0.00019873298478639725,
+      "loss": 2.3931,
+      "step": 270
+    },
+    {
+      "epoch": 0.006594131223211342,
+      "grad_norm": 1.6674144268035889,
+      "learning_rate": 0.00019868588384908863,
+      "loss": 2.6797,
+      "step": 280
+    },
+    {
+      "epoch": 0.006829635909754604,
+      "grad_norm": 1.8776614665985107,
+      "learning_rate": 0.00019863878291177995,
+      "loss": 2.3755,
+      "step": 290
+    },
+    {
+      "epoch": 0.007065140596297866,
+      "grad_norm": 1.704877257347107,
+      "learning_rate": 0.0001985916819744713,
+      "loss": 2.577,
+      "step": 300
+    },
+    {
+      "epoch": 0.007300645282841129,
+      "grad_norm": 1.7447631359100342,
+      "learning_rate": 0.00019854458103716265,
+      "loss": 2.6707,
+      "step": 310
+    },
+    {
+      "epoch": 0.007536149969384391,
+      "grad_norm": 1.7486447095870972,
+      "learning_rate": 0.000198497480099854,
+      "loss": 2.4179,
+      "step": 320
+    },
+    {
+      "epoch": 0.007771654655927653,
+      "grad_norm": 1.5924261808395386,
+      "learning_rate": 0.00019845037916254535,
+      "loss": 2.3293,
+      "step": 330
+    },
+    {
+      "epoch": 0.008007159342470916,
+      "grad_norm": 2.1535685062408447,
+      "learning_rate": 0.0001984032782252367,
+      "loss": 2.4561,
+      "step": 340
+    },
+    {
+      "epoch": 0.008242664029014177,
+      "grad_norm": 1.8906278610229492,
+      "learning_rate": 0.00019835617728792803,
+      "loss": 2.5716,
+      "step": 350
+    },
+    {
+      "epoch": 0.00847816871555744,
+      "grad_norm": 2.237260580062866,
+      "learning_rate": 0.0001983090763506194,
+      "loss": 2.4573,
+      "step": 360
+    },
+    {
+      "epoch": 0.008713673402100702,
+      "grad_norm": 1.506896734237671,
+      "learning_rate": 0.00019826197541331073,
+      "loss": 2.4239,
+      "step": 370
+    },
+    {
+      "epoch": 0.008949178088643964,
+      "grad_norm": 1.8521124124526978,
+      "learning_rate": 0.00019821487447600208,
+      "loss": 2.518,
+      "step": 380
+    },
+    {
+      "epoch": 0.009184682775187226,
+      "grad_norm": 1.684022307395935,
+      "learning_rate": 0.00019816777353869343,
+      "loss": 2.2849,
+      "step": 390
+    },
+    {
+      "epoch": 0.009420187461730489,
+      "grad_norm": 1.6935802698135376,
+      "learning_rate": 0.00019812067260138478,
+      "loss": 2.5127,
+      "step": 400
+    },
+    {
+      "epoch": 0.009655692148273751,
+      "grad_norm": 2.0532476902008057,
+      "learning_rate": 0.00019807357166407613,
+      "loss": 2.2883,
+      "step": 410
+    },
+    {
+      "epoch": 0.009891196834817012,
+      "grad_norm": 1.579057216644287,
+      "learning_rate": 0.00019802647072676748,
+      "loss": 2.3682,
+      "step": 420
+    },
+    {
+      "epoch": 0.010126701521360275,
+      "grad_norm": 1.471557855606079,
+      "learning_rate": 0.0001979793697894588,
+      "loss": 2.4559,
+      "step": 430
+    },
+    {
+      "epoch": 0.010362206207903537,
+      "grad_norm": 1.4747744798660278,
+      "learning_rate": 0.00019793226885215018,
+      "loss": 2.3266,
+      "step": 440
+    },
+    {
+      "epoch": 0.0105977108944468,
+      "grad_norm": 1.7954132556915283,
+      "learning_rate": 0.0001978851679148415,
+      "loss": 2.4466,
+      "step": 450
+    },
+    {
+      "epoch": 0.010833215580990062,
+      "grad_norm": 1.8787152767181396,
+      "learning_rate": 0.00019783806697753285,
+      "loss": 2.2744,
+      "step": 460
+    },
+    {
+      "epoch": 0.011068720267533324,
+      "grad_norm": 2.769911527633667,
+      "learning_rate": 0.0001977909660402242,
+      "loss": 2.4138,
+      "step": 470
+    },
+    {
+      "epoch": 0.011304224954076587,
+      "grad_norm": 1.7480900287628174,
+      "learning_rate": 0.00019774386510291555,
+      "loss": 2.3967,
+      "step": 480
+    },
+    {
+      "epoch": 0.01153972964061985,
+      "grad_norm": 1.5646064281463623,
+      "learning_rate": 0.0001976967641656069,
+      "loss": 2.3035,
+      "step": 490
+    },
+    {
+      "epoch": 0.01177523432716311,
+      "grad_norm": 2.1122241020202637,
+      "learning_rate": 0.00019764966322829825,
+      "loss": 2.4634,
+      "step": 500
+    },
+    {
+      "epoch": 0.012010739013706372,
+      "grad_norm": 1.5615967512130737,
+      "learning_rate": 0.00019760256229098958,
+      "loss": 2.4727,
+      "step": 510
+    },
+    {
+      "epoch": 0.012246243700249635,
+      "grad_norm": 1.6492537260055542,
+      "learning_rate": 0.00019755546135368095,
+      "loss": 2.309,
+      "step": 520
+    },
+    {
+      "epoch": 0.012481748386792897,
+      "grad_norm": 1.8484909534454346,
+      "learning_rate": 0.00019750836041637228,
+      "loss": 2.5138,
+      "step": 530
+    },
+    {
+      "epoch": 0.01271725307333616,
+      "grad_norm": 1.2929155826568604,
+      "learning_rate": 0.00019746125947906365,
+      "loss": 2.1399,
+      "step": 540
+    },
+    {
+      "epoch": 0.012952757759879422,
+      "grad_norm": 2.6417770385742188,
+      "learning_rate": 0.000197414158541755,
+      "loss": 2.5062,
+      "step": 550
+    },
+    {
+      "epoch": 0.013188262446422684,
+      "grad_norm": 1.4375731945037842,
+      "learning_rate": 0.00019736705760444633,
+      "loss": 2.218,
+      "step": 560
+    },
+    {
+      "epoch": 0.013423767132965945,
+      "grad_norm": 1.6705974340438843,
+      "learning_rate": 0.0001973199566671377,
+      "loss": 2.1678,
+      "step": 570
+    },
+    {
+      "epoch": 0.013659271819509208,
+      "grad_norm": 1.7592695951461792,
+      "learning_rate": 0.00019727285572982903,
+      "loss": 2.1822,
+      "step": 580
+    },
+    {
+      "epoch": 0.01389477650605247,
+      "grad_norm": 1.7886682748794556,
+      "learning_rate": 0.00019722575479252038,
+      "loss": 2.3795,
+      "step": 590
+    },
+    {
+      "epoch": 0.014130281192595733,
+      "grad_norm": 1.9300228357315063,
+      "learning_rate": 0.00019717865385521173,
+      "loss": 2.3488,
+      "step": 600
+    },
+    {
+      "epoch": 0.014365785879138995,
+      "grad_norm": 1.589822769165039,
+      "learning_rate": 0.00019713155291790308,
+      "loss": 2.4307,
+      "step": 610
+    },
+    {
+      "epoch": 0.014601290565682257,
+      "grad_norm": 1.503128170967102,
+      "learning_rate": 0.00019708445198059443,
+      "loss": 2.277,
+      "step": 620
+    },
+    {
+      "epoch": 0.01483679525222552,
+      "grad_norm": 1.7925318479537964,
+      "learning_rate": 0.00019703735104328578,
+      "loss": 2.3734,
+      "step": 630
+    },
+    {
+      "epoch": 0.015072299938768782,
+      "grad_norm": 1.6709870100021362,
+      "learning_rate": 0.0001969902501059771,
+      "loss": 2.1709,
+      "step": 640
+    },
+    {
+      "epoch": 0.015307804625312043,
+      "grad_norm": 1.6741284132003784,
+      "learning_rate": 0.00019694314916866848,
+      "loss": 2.2209,
+      "step": 650
+    },
+    {
+      "epoch": 0.015543309311855305,
+      "grad_norm": 1.6825313568115234,
+      "learning_rate": 0.0001968960482313598,
+      "loss": 2.255,
+      "step": 660
+    },
+    {
+      "epoch": 0.015778813998398568,
+      "grad_norm": 1.544534683227539,
+      "learning_rate": 0.00019684894729405115,
+      "loss": 2.3872,
+      "step": 670
+    },
+    {
+      "epoch": 0.016014318684941832,
+      "grad_norm": 2.33375883102417,
+      "learning_rate": 0.0001968018463567425,
+      "loss": 2.4629,
+      "step": 680
+    },
+    {
+      "epoch": 0.016249823371485093,
+      "grad_norm": 1.4496179819107056,
+      "learning_rate": 0.00019675474541943385,
+      "loss": 2.3638,
+      "step": 690
+    },
+    {
+      "epoch": 0.016485328058028353,
+      "grad_norm": 1.7347285747528076,
+      "learning_rate": 0.0001967076444821252,
+      "loss": 2.4611,
+      "step": 700
+    },
+    {
+      "epoch": 0.016720832744571618,
+      "grad_norm": 1.5036282539367676,
+      "learning_rate": 0.00019666054354481655,
+      "loss": 2.3117,
+      "step": 710
+    },
+    {
+      "epoch": 0.01695633743111488,
+      "grad_norm": 2.0738086700439453,
+      "learning_rate": 0.0001966134426075079,
+      "loss": 2.3374,
+      "step": 720
+    },
+    {
+      "epoch": 0.017191842117658142,
+      "grad_norm": 2.5572900772094727,
+      "learning_rate": 0.00019656634167019926,
+      "loss": 2.3246,
+      "step": 730
+    },
+    {
+      "epoch": 0.017427346804201403,
+      "grad_norm": 1.6567786931991577,
+      "learning_rate": 0.00019651924073289058,
+      "loss": 2.4771,
+      "step": 740
+    },
+    {
+      "epoch": 0.017662851490744667,
+      "grad_norm": 2.6085476875305176,
+      "learning_rate": 0.00019647213979558196,
+      "loss": 2.3132,
+      "step": 750
+    },
+    {
+      "epoch": 0.017898356177287928,
+      "grad_norm": 1.939334511756897,
+      "learning_rate": 0.00019642503885827328,
+      "loss": 2.4239,
+      "step": 760
+    },
+    {
+      "epoch": 0.01813386086383119,
+      "grad_norm": 1.840398907661438,
+      "learning_rate": 0.00019637793792096463,
+      "loss": 2.3915,
+      "step": 770
+    },
+    {
+      "epoch": 0.018369365550374453,
+      "grad_norm": 1.8545604944229126,
+      "learning_rate": 0.00019633083698365598,
+      "loss": 2.3342,
+      "step": 780
+    },
+    {
+      "epoch": 0.018604870236917714,
+      "grad_norm": 1.8592214584350586,
+      "learning_rate": 0.00019628373604634733,
+      "loss": 2.4354,
+      "step": 790
+    },
+    {
+      "epoch": 0.018840374923460978,
+      "grad_norm": 1.6031461954116821,
+      "learning_rate": 0.00019623663510903868,
+      "loss": 2.3723,
+      "step": 800
+    },
+    {
+      "epoch": 0.01907587961000424,
+      "grad_norm": 2.243943214416504,
+      "learning_rate": 0.00019618953417173003,
+      "loss": 2.5432,
+      "step": 810
+    },
+    {
+      "epoch": 0.019311384296547503,
+      "grad_norm": 1.997883677482605,
+      "learning_rate": 0.00019614243323442135,
+      "loss": 2.2411,
+      "step": 820
+    },
+    {
+      "epoch": 0.019546888983090763,
+      "grad_norm": 1.7321577072143555,
+      "learning_rate": 0.00019609533229711273,
+      "loss": 2.2699,
+      "step": 830
+    },
+    {
+      "epoch": 0.019782393669634024,
+      "grad_norm": 1.544219732284546,
+      "learning_rate": 0.00019604823135980408,
+      "loss": 2.265,
+      "step": 840
+    },
+    {
+      "epoch": 0.02001789835617729,
+      "grad_norm": 1.6504725217819214,
+      "learning_rate": 0.0001960011304224954,
+      "loss": 2.4386,
+      "step": 850
+    },
+    {
+      "epoch": 0.02025340304272055,
+      "grad_norm": 2.053715705871582,
+      "learning_rate": 0.00019595402948518678,
+      "loss": 2.2794,
+      "step": 860
+    },
+    {
+      "epoch": 0.020488907729263813,
+      "grad_norm": 1.469041347503662,
+      "learning_rate": 0.0001959069285478781,
+      "loss": 2.4016,
+      "step": 870
+    },
+    {
+      "epoch": 0.020724412415807074,
+      "grad_norm": 1.7059731483459473,
+      "learning_rate": 0.00019585982761056948,
+      "loss": 2.3508,
+      "step": 880
+    },
+    {
+      "epoch": 0.020959917102350338,
+      "grad_norm": 1.9896105527877808,
+      "learning_rate": 0.0001958127266732608,
+      "loss": 2.3903,
+      "step": 890
+    },
+    {
+      "epoch": 0.0211954217888936,
+      "grad_norm": 2.041901111602783,
+      "learning_rate": 0.00019576562573595216,
+      "loss": 2.4199,
+      "step": 900
+    },
+    {
+      "epoch": 0.02143092647543686,
+      "grad_norm": 1.4538580179214478,
+      "learning_rate": 0.0001957185247986435,
+      "loss": 2.2481,
+      "step": 910
+    },
+    {
+      "epoch": 0.021666431161980124,
+      "grad_norm": 2.297203302383423,
+      "learning_rate": 0.00019567142386133486,
+      "loss": 2.3375,
+      "step": 920
+    },
+    {
+      "epoch": 0.021901935848523384,
+      "grad_norm": 2.172999143600464,
+      "learning_rate": 0.0001956243229240262,
+      "loss": 2.223,
+      "step": 930
+    },
+    {
+      "epoch": 0.02213744053506665,
+      "grad_norm": 1.5507631301879883,
+      "learning_rate": 0.00019557722198671756,
+      "loss": 2.2435,
+      "step": 940
+    },
+    {
+      "epoch": 0.02237294522160991,
+      "grad_norm": 2.088479518890381,
+      "learning_rate": 0.00019553012104940888,
+      "loss": 2.1769,
+      "step": 950
+    },
+    {
+      "epoch": 0.022608449908153173,
+      "grad_norm": 1.3809598684310913,
+      "learning_rate": 0.00019548302011210026,
+      "loss": 2.3431,
+      "step": 960
+    },
+    {
+      "epoch": 0.022843954594696434,
+      "grad_norm": 1.5656312704086304,
+      "learning_rate": 0.00019543591917479158,
+      "loss": 2.2484,
+      "step": 970
+    },
+    {
+      "epoch": 0.0230794592812397,
+      "grad_norm": 1.9035502672195435,
+      "learning_rate": 0.00019538881823748293,
+      "loss": 2.1907,
+      "step": 980
+    },
+    {
+      "epoch": 0.02331496396778296,
+      "grad_norm": 2.3373143672943115,
+      "learning_rate": 0.00019534171730017428,
+      "loss": 2.2871,
+      "step": 990
+    },
+    {
+      "epoch": 0.02355046865432622,
+      "grad_norm": 1.9408587217330933,
+      "learning_rate": 0.00019529461636286563,
+      "loss": 1.9912,
+      "step": 1000
+    },
+    {
+      "epoch": 0.023785973340869484,
+      "grad_norm": 2.7549495697021484,
+      "learning_rate": 0.00019524751542555698,
+      "loss": 2.4807,
+      "step": 1010
+    },
+    {
+      "epoch": 0.024021478027412745,
+      "grad_norm": 1.6193737983703613,
+      "learning_rate": 0.00019520041448824833,
+      "loss": 2.2551,
+      "step": 1020
+    },
+    {
+      "epoch": 0.02425698271395601,
+      "grad_norm": 2.092519760131836,
+      "learning_rate": 0.00019515331355093965,
+      "loss": 2.0965,
+      "step": 1030
+    },
+    {
+      "epoch": 0.02449248740049927,
+      "grad_norm": 2.3660218715667725,
+      "learning_rate": 0.00019510621261363103,
+      "loss": 2.3436,
+      "step": 1040
+    },
+    {
+      "epoch": 0.024727992087042534,
+      "grad_norm": 2.076950788497925,
+      "learning_rate": 0.00019505911167632235,
+      "loss": 2.0685,
+      "step": 1050
+    },
+    {
+      "epoch": 0.024963496773585794,
+      "grad_norm": 1.881016731262207,
+      "learning_rate": 0.0001950120107390137,
+      "loss": 2.524,
+      "step": 1060
+    },
+    {
+      "epoch": 0.025199001460129055,
+      "grad_norm": 1.6038298606872559,
+      "learning_rate": 0.00019496490980170506,
+      "loss": 2.3772,
+      "step": 1070
+    },
+    {
+      "epoch": 0.02543450614667232,
+      "grad_norm": 1.5274938344955444,
+      "learning_rate": 0.0001949178088643964,
+      "loss": 2.3456,
+      "step": 1080
+    },
+    {
+      "epoch": 0.02567001083321558,
+      "grad_norm": 1.5426788330078125,
+      "learning_rate": 0.00019487070792708776,
+      "loss": 2.4562,
+      "step": 1090
+    },
+    {
+      "epoch": 0.025905515519758844,
+      "grad_norm": 1.9040606021881104,
+      "learning_rate": 0.0001948236069897791,
+      "loss": 2.2722,
+      "step": 1100
+    },
+    {
+      "epoch": 0.026141020206302105,
+      "grad_norm": 2.3787999153137207,
+      "learning_rate": 0.00019477650605247046,
+      "loss": 2.4295,
+      "step": 1110
+    },
+    {
+      "epoch": 0.02637652489284537,
+      "grad_norm": 1.6062979698181152,
+      "learning_rate": 0.0001947294051151618,
+      "loss": 2.3269,
+      "step": 1120
+    },
+    {
+      "epoch": 0.02661202957938863,
+      "grad_norm": 1.6376535892486572,
+      "learning_rate": 0.00019468230417785316,
+      "loss": 2.3202,
+      "step": 1130
+    },
+    {
+      "epoch": 0.02684753426593189,
+      "grad_norm": 1.78897225856781,
+      "learning_rate": 0.0001946352032405445,
+      "loss": 2.4023,
+      "step": 1140
+    },
+    {
+      "epoch": 0.027083038952475155,
+      "grad_norm": 1.784119963645935,
+      "learning_rate": 0.00019458810230323586,
+      "loss": 2.177,
+      "step": 1150
+    },
+    {
+      "epoch": 0.027318543639018415,
+      "grad_norm": 2.4228804111480713,
+      "learning_rate": 0.00019454100136592718,
+      "loss": 2.4059,
+      "step": 1160
+    },
+    {
+      "epoch": 0.02755404832556168,
+      "grad_norm": 2.120675802230835,
+      "learning_rate": 0.00019449390042861856,
+      "loss": 2.2732,
+      "step": 1170
+    },
+    {
+      "epoch": 0.02778955301210494,
+      "grad_norm": 1.370073676109314,
+      "learning_rate": 0.00019444679949130988,
+      "loss": 2.2269,
+      "step": 1180
+    },
+    {
+      "epoch": 0.028025057698648204,
+      "grad_norm": 1.6620279550552368,
+      "learning_rate": 0.00019439969855400123,
+      "loss": 2.2124,
+      "step": 1190
+    },
+    {
+      "epoch": 0.028260562385191465,
+      "grad_norm": 2.0860698223114014,
+      "learning_rate": 0.00019435259761669258,
+      "loss": 2.1872,
+      "step": 1200
+    },
+    {
+      "epoch": 0.02849606707173473,
+      "grad_norm": 1.944350004196167,
+      "learning_rate": 0.00019430549667938393,
+      "loss": 2.3289,
+      "step": 1210
+    },
+    {
+      "epoch": 0.02873157175827799,
+      "grad_norm": 1.8327244520187378,
+      "learning_rate": 0.00019425839574207528,
+      "loss": 2.2747,
+      "step": 1220
+    },
+    {
+      "epoch": 0.02896707644482125,
+      "grad_norm": 1.52091383934021,
+      "learning_rate": 0.00019421129480476663,
+      "loss": 2.3557,
+      "step": 1230
+    },
+    {
+      "epoch": 0.029202581131364515,
+      "grad_norm": 1.7364777326583862,
+      "learning_rate": 0.00019416419386745796,
+      "loss": 2.2268,
+      "step": 1240
+    },
+    {
+      "epoch": 0.029438085817907775,
+      "grad_norm": 1.7765967845916748,
+      "learning_rate": 0.00019411709293014933,
+      "loss": 2.1311,
+      "step": 1250
+    },
+    {
+      "epoch": 0.02967359050445104,
+      "grad_norm": 2.070098876953125,
+      "learning_rate": 0.00019406999199284066,
+      "loss": 2.2586,
+      "step": 1260
+    },
+    {
+      "epoch": 0.0299090951909943,
+      "grad_norm": 1.7408677339553833,
+      "learning_rate": 0.000194022891055532,
+      "loss": 2.2579,
+      "step": 1270
+    },
+    {
+      "epoch": 0.030144599877537565,
+      "grad_norm": 1.849951982498169,
+      "learning_rate": 0.00019397579011822336,
+      "loss": 2.3557,
+      "step": 1280
+    },
+    {
+      "epoch": 0.030380104564080825,
+      "grad_norm": 1.4832192659378052,
+      "learning_rate": 0.0001939286891809147,
+      "loss": 2.0548,
+      "step": 1290
+    },
+    {
+      "epoch": 0.030615609250624086,
+      "grad_norm": 2.5369913578033447,
+      "learning_rate": 0.00019388158824360606,
+      "loss": 2.2137,
+      "step": 1300
+    },
+    {
+      "epoch": 0.03085111393716735,
+      "grad_norm": 1.648578405380249,
+      "learning_rate": 0.0001938344873062974,
+      "loss": 2.4065,
+      "step": 1310
+    },
+    {
+      "epoch": 0.03108661862371061,
+      "grad_norm": 1.6784148216247559,
+      "learning_rate": 0.00019378738636898876,
+      "loss": 2.1999,
+      "step": 1320
+    },
+    {
+      "epoch": 0.03132212331025387,
+      "grad_norm": 2.132889747619629,
+      "learning_rate": 0.0001937402854316801,
+      "loss": 2.3155,
+      "step": 1330
+    },
+    {
+      "epoch": 0.031557627996797136,
+      "grad_norm": 1.9378546476364136,
+      "learning_rate": 0.00019369318449437143,
+      "loss": 2.2407,
+      "step": 1340
+    },
+    {
+      "epoch": 0.0317931326833404,
+      "grad_norm": 4.045490741729736,
+      "learning_rate": 0.0001936460835570628,
+      "loss": 2.2026,
+      "step": 1350
+    },
+    {
+      "epoch": 0.032028637369883664,
+      "grad_norm": 1.7972216606140137,
+      "learning_rate": 0.00019359898261975413,
+      "loss": 2.3747,
+      "step": 1360
+    },
+    {
+      "epoch": 0.03226414205642692,
+      "grad_norm": 1.7621222734451294,
+      "learning_rate": 0.00019355188168244548,
+      "loss": 2.1652,
+      "step": 1370
+    },
+    {
+      "epoch": 0.032499646742970185,
+      "grad_norm": 2.10644268989563,
+      "learning_rate": 0.00019350478074513683,
+      "loss": 2.4651,
+      "step": 1380
+    },
+    {
+      "epoch": 0.03273515142951345,
+      "grad_norm": 1.6137688159942627,
+      "learning_rate": 0.00019345767980782818,
+      "loss": 2.1893,
+      "step": 1390
+    },
+    {
+      "epoch": 0.03297065611605671,
+      "grad_norm": 2.142392873764038,
+      "learning_rate": 0.00019341057887051953,
+      "loss": 2.3385,
+      "step": 1400
+    },
+    {
+      "epoch": 0.03320616080259997,
+      "grad_norm": 2.0509915351867676,
+      "learning_rate": 0.00019336347793321088,
+      "loss": 2.2966,
+      "step": 1410
+    },
+    {
+      "epoch": 0.033441665489143235,
+      "grad_norm": 1.7054039239883423,
+      "learning_rate": 0.00019331637699590223,
+      "loss": 2.2535,
+      "step": 1420
+    },
+    {
+      "epoch": 0.0336771701756865,
+      "grad_norm": 1.9987514019012451,
+      "learning_rate": 0.00019326927605859358,
+      "loss": 2.2129,
+      "step": 1430
+    },
+    {
+      "epoch": 0.03391267486222976,
+      "grad_norm": 1.558950424194336,
+      "learning_rate": 0.00019322217512128493,
+      "loss": 2.3284,
+      "step": 1440
+    },
+    {
+      "epoch": 0.03414817954877302,
+      "grad_norm": 2.1738619804382324,
+      "learning_rate": 0.00019317507418397626,
+      "loss": 2.2122,
+      "step": 1450
+    },
+    {
+      "epoch": 0.034383684235316285,
+      "grad_norm": 2.3210151195526123,
+      "learning_rate": 0.00019312797324666763,
+      "loss": 2.1787,
+      "step": 1460
+    },
+    {
+      "epoch": 0.03461918892185954,
+      "grad_norm": 2.7615976333618164,
+      "learning_rate": 0.00019308087230935896,
+      "loss": 2.1883,
+      "step": 1470
+    },
+    {
+      "epoch": 0.034854693608402806,
+      "grad_norm": 1.751855731010437,
+      "learning_rate": 0.00019303377137205033,
+      "loss": 2.2706,
+      "step": 1480
+    },
+    {
+      "epoch": 0.03509019829494607,
+      "grad_norm": 1.46992027759552,
+      "learning_rate": 0.00019298667043474166,
+      "loss": 2.1759,
+      "step": 1490
+    },
+    {
+      "epoch": 0.035325702981489335,
+      "grad_norm": 1.7911906242370605,
+      "learning_rate": 0.000192939569497433,
+      "loss": 2.0725,
+      "step": 1500
+    },
+    {
+      "epoch": 0.03556120766803259,
+      "grad_norm": 1.5120116472244263,
+      "learning_rate": 0.00019289246856012436,
+      "loss": 2.3206,
+      "step": 1510
+    },
+    {
+      "epoch": 0.035796712354575856,
+      "grad_norm": 1.7404943704605103,
+      "learning_rate": 0.0001928453676228157,
+      "loss": 2.1382,
+      "step": 1520
+    },
+    {
+      "epoch": 0.03603221704111912,
+      "grad_norm": 1.7643208503723145,
+      "learning_rate": 0.00019279826668550706,
+      "loss": 2.0674,
+      "step": 1530
+    },
+    {
+      "epoch": 0.03626772172766238,
+      "grad_norm": 2.0491485595703125,
+      "learning_rate": 0.0001927511657481984,
+      "loss": 2.378,
+      "step": 1540
+    },
+    {
+      "epoch": 0.03650322641420564,
+      "grad_norm": 1.71403169631958,
+      "learning_rate": 0.00019270406481088973,
+      "loss": 2.3117,
+      "step": 1550
+    },
+    {
+      "epoch": 0.036738731100748906,
+      "grad_norm": 1.729164719581604,
+      "learning_rate": 0.0001926569638735811,
+      "loss": 2.2056,
+      "step": 1560
+    },
+    {
+      "epoch": 0.03697423578729217,
+      "grad_norm": 2.1834709644317627,
+      "learning_rate": 0.00019260986293627243,
+      "loss": 2.265,
+      "step": 1570
+    },
+    {
+      "epoch": 0.03720974047383543,
+      "grad_norm": 1.5792131423950195,
+      "learning_rate": 0.00019256276199896378,
+      "loss": 2.2569,
+      "step": 1580
+    },
+    {
+      "epoch": 0.03744524516037869,
+      "grad_norm": 2.3097856044769287,
+      "learning_rate": 0.00019251566106165513,
+      "loss": 2.2394,
+      "step": 1590
+    },
+    {
+      "epoch": 0.037680749846921956,
+      "grad_norm": 1.9641480445861816,
+      "learning_rate": 0.00019246856012434648,
+      "loss": 2.3833,
+      "step": 1600
+    },
+    {
+      "epoch": 0.03791625453346521,
+      "grad_norm": 2.1574530601501465,
+      "learning_rate": 0.00019242145918703783,
+      "loss": 2.1989,
+      "step": 1610
+    },
+    {
+      "epoch": 0.03815175922000848,
+      "grad_norm": 2.2117180824279785,
+      "learning_rate": 0.00019237435824972918,
+      "loss": 2.286,
+      "step": 1620
+    },
+    {
+      "epoch": 0.03838726390655174,
+      "grad_norm": 1.692057728767395,
+      "learning_rate": 0.0001923272573124205,
+      "loss": 2.3328,
+      "step": 1630
+    },
+    {
+      "epoch": 0.038622768593095005,
+      "grad_norm": 2.3093721866607666,
+      "learning_rate": 0.00019228015637511188,
+      "loss": 2.284,
+      "step": 1640
+    },
+    {
+      "epoch": 0.03885827327963826,
+      "grad_norm": 1.6981611251831055,
+      "learning_rate": 0.0001922330554378032,
+      "loss": 2.1517,
+      "step": 1650
+    },
+    {
+      "epoch": 0.03909377796618153,
+      "grad_norm": 2.144850730895996,
+      "learning_rate": 0.00019218595450049456,
+      "loss": 2.3897,
+      "step": 1660
+    },
+    {
+      "epoch": 0.03932928265272479,
+      "grad_norm": 1.5663037300109863,
+      "learning_rate": 0.00019213885356318593,
+      "loss": 2.2038,
+      "step": 1670
+    },
+    {
+      "epoch": 0.03956478733926805,
+      "grad_norm": 2.053711175918579,
+      "learning_rate": 0.00019209175262587726,
+      "loss": 2.1427,
+      "step": 1680
+    },
+    {
+      "epoch": 0.03980029202581131,
+      "grad_norm": 1.7935670614242554,
+      "learning_rate": 0.00019204465168856863,
+      "loss": 2.3839,
+      "step": 1690
+    },
+    {
+      "epoch": 0.04003579671235458,
+      "grad_norm": 2.1188805103302,
+      "learning_rate": 0.00019199755075125996,
+      "loss": 2.2879,
+      "step": 1700
+    },
+    {
+      "epoch": 0.04027130139889784,
+      "grad_norm": 1.8197453022003174,
+      "learning_rate": 0.0001919504498139513,
+      "loss": 2.361,
+      "step": 1710
+    },
+    {
+      "epoch": 0.0405068060854411,
+      "grad_norm": 1.9226861000061035,
+      "learning_rate": 0.00019190334887664266,
+      "loss": 2.3839,
+      "step": 1720
+    },
+    {
+      "epoch": 0.04074231077198436,
+      "grad_norm": 1.8571581840515137,
+      "learning_rate": 0.000191856247939334,
+      "loss": 2.2876,
+      "step": 1730
+    },
+    {
+      "epoch": 0.040977815458527626,
+      "grad_norm": 1.9353306293487549,
+      "learning_rate": 0.00019180914700202536,
+      "loss": 2.1383,
+      "step": 1740
+    },
+    {
+      "epoch": 0.041213320145070884,
+      "grad_norm": 2.7373709678649902,
+      "learning_rate": 0.0001917620460647167,
+      "loss": 2.1032,
+      "step": 1750
+    },
+    {
+      "epoch": 0.04144882483161415,
+      "grad_norm": 1.6866846084594727,
+      "learning_rate": 0.00019171494512740803,
+      "loss": 2.3656,
+      "step": 1760
+    },
+    {
+      "epoch": 0.04168432951815741,
+      "grad_norm": 1.7577126026153564,
+      "learning_rate": 0.0001916678441900994,
+      "loss": 2.2786,
+      "step": 1770
+    },
+    {
+      "epoch": 0.041919834204700676,
+      "grad_norm": 2.247462272644043,
+      "learning_rate": 0.00019162074325279073,
+      "loss": 2.252,
+      "step": 1780
+    },
+    {
+      "epoch": 0.04215533889124393,
+      "grad_norm": 1.7799142599105835,
+      "learning_rate": 0.00019157364231548208,
+      "loss": 2.2314,
+      "step": 1790
+    },
+    {
+      "epoch": 0.0423908435777872,
+      "grad_norm": 2.1439526081085205,
+      "learning_rate": 0.00019152654137817343,
+      "loss": 2.4094,
+      "step": 1800
+    },
+    {
+      "epoch": 0.04262634826433046,
+      "grad_norm": 2.395840644836426,
+      "learning_rate": 0.00019147944044086478,
+      "loss": 2.1436,
+      "step": 1810
+    },
+    {
+      "epoch": 0.04286185295087372,
+      "grad_norm": 1.6308691501617432,
+      "learning_rate": 0.00019143233950355613,
+      "loss": 2.2567,
+      "step": 1820
+    },
+    {
+      "epoch": 0.04309735763741698,
+      "grad_norm": 2.162234306335449,
+      "learning_rate": 0.00019138523856624748,
+      "loss": 2.3629,
+      "step": 1830
+    },
+    {
+      "epoch": 0.04333286232396025,
+      "grad_norm": 2.5992627143859863,
+      "learning_rate": 0.0001913381376289388,
+      "loss": 2.192,
+      "step": 1840
+    },
+    {
+      "epoch": 0.04356836701050351,
+      "grad_norm": 1.7631303071975708,
+      "learning_rate": 0.00019129103669163018,
+      "loss": 2.2224,
+      "step": 1850
+    },
+    {
+      "epoch": 0.04380387169704677,
+      "grad_norm": 1.678625464439392,
+      "learning_rate": 0.0001912439357543215,
+      "loss": 2.2322,
+      "step": 1860
+    },
+    {
+      "epoch": 0.04403937638359003,
+      "grad_norm": 2.3097527027130127,
+      "learning_rate": 0.00019119683481701286,
+      "loss": 2.3262,
+      "step": 1870
+    },
+    {
+      "epoch": 0.0442748810701333,
+      "grad_norm": 1.3386963605880737,
+      "learning_rate": 0.0001911497338797042,
+      "loss": 2.1346,
+      "step": 1880
+    },
+    {
+      "epoch": 0.04451038575667656,
+      "grad_norm": 2.4035725593566895,
+      "learning_rate": 0.00019110263294239556,
+      "loss": 2.2836,
+      "step": 1890
+    },
+    {
+      "epoch": 0.04474589044321982,
+      "grad_norm": 1.9946550130844116,
+      "learning_rate": 0.0001910555320050869,
+      "loss": 2.2045,
+      "step": 1900
+    },
+    {
+      "epoch": 0.04498139512976308,
+      "grad_norm": 1.757352590560913,
+      "learning_rate": 0.00019100843106777826,
+      "loss": 1.9896,
+      "step": 1910
+    },
+    {
+      "epoch": 0.04521689981630635,
+      "grad_norm": 1.8178822994232178,
+      "learning_rate": 0.0001909613301304696,
+      "loss": 2.2243,
+      "step": 1920
+    },
+    {
+      "epoch": 0.045452404502849604,
+      "grad_norm": 2.1335227489471436,
+      "learning_rate": 0.00019091422919316096,
+      "loss": 2.4226,
+      "step": 1930
+    },
+    {
+      "epoch": 0.04568790918939287,
+      "grad_norm": 2.0507328510284424,
+      "learning_rate": 0.00019086712825585228,
+      "loss": 2.3667,
+      "step": 1940
+    },
+    {
+      "epoch": 0.04592341387593613,
+      "grad_norm": 2.5613133907318115,
+      "learning_rate": 0.00019082002731854366,
+      "loss": 2.4204,
+      "step": 1950
+    },
+    {
+      "epoch": 0.0461589185624794,
+      "grad_norm": 2.1812310218811035,
+      "learning_rate": 0.000190772926381235,
+      "loss": 2.1989,
+      "step": 1960
+    },
+    {
+      "epoch": 0.046394423249022654,
+      "grad_norm": 2.316821336746216,
+      "learning_rate": 0.00019072582544392633,
+      "loss": 2.4335,
+      "step": 1970
+    },
+    {
+      "epoch": 0.04662992793556592,
+      "grad_norm": 2.272609233856201,
+      "learning_rate": 0.0001906787245066177,
+      "loss": 2.2962,
+      "step": 1980
+    },
+    {
+      "epoch": 0.04686543262210918,
+      "grad_norm": 2.0263876914978027,
+      "learning_rate": 0.00019063162356930903,
+      "loss": 2.2607,
+      "step": 1990
+    },
+    {
+      "epoch": 0.04710093730865244,
+      "grad_norm": 2.9174461364746094,
+      "learning_rate": 0.00019058452263200038,
+      "loss": 2.2248,
+      "step": 2000
+    },
+    {
+      "epoch": 0.047336441995195704,
+      "grad_norm": 2.3031280040740967,
+      "learning_rate": 0.00019053742169469173,
+      "loss": 2.3064,
+      "step": 2010
+    },
+    {
+      "epoch": 0.04757194668173897,
+      "grad_norm": 1.9068163633346558,
+      "learning_rate": 0.00019049032075738308,
+      "loss": 2.1707,
+      "step": 2020
+    },
+    {
+      "epoch": 0.04780745136828223,
+      "grad_norm": 1.6343876123428345,
+      "learning_rate": 0.00019044321982007443,
+      "loss": 2.1399,
+      "step": 2030
+    },
+    {
+      "epoch": 0.04804295605482549,
+      "grad_norm": 2.0931334495544434,
+      "learning_rate": 0.00019039611888276578,
+      "loss": 2.1365,
+      "step": 2040
+    },
+    {
+      "epoch": 0.04827846074136875,
+      "grad_norm": 2.3466007709503174,
+      "learning_rate": 0.0001903490179454571,
+      "loss": 2.2243,
+      "step": 2050
+    },
+    {
+      "epoch": 0.04851396542791202,
+      "grad_norm": 1.7455843687057495,
+      "learning_rate": 0.00019030191700814849,
+      "loss": 2.2342,
+      "step": 2060
+    },
+    {
+      "epoch": 0.048749470114455275,
+      "grad_norm": 2.6780898571014404,
+      "learning_rate": 0.0001902548160708398,
+      "loss": 2.3248,
+      "step": 2070
+    },
+    {
+      "epoch": 0.04898497480099854,
+      "grad_norm": 2.0855660438537598,
+      "learning_rate": 0.00019020771513353119,
+      "loss": 2.3911,
+      "step": 2080
+    },
+    {
+      "epoch": 0.0492204794875418,
+      "grad_norm": 1.9923450946807861,
+      "learning_rate": 0.0001901606141962225,
+      "loss": 2.2859,
+      "step": 2090
+    },
+    {
+      "epoch": 0.04945598417408507,
+      "grad_norm": 1.83405601978302,
+      "learning_rate": 0.00019011351325891386,
+      "loss": 2.0618,
+      "step": 2100
+    },
+    {
+      "epoch": 0.049691488860628324,
+      "grad_norm": 2.5639121532440186,
+      "learning_rate": 0.0001900664123216052,
+      "loss": 2.3821,
+      "step": 2110
+    },
+    {
+      "epoch": 0.04992699354717159,
+      "grad_norm": 1.992323875427246,
+      "learning_rate": 0.00019001931138429656,
+      "loss": 2.2083,
+      "step": 2120
+    },
+    {
+      "epoch": 0.05016249823371485,
+      "grad_norm": 2.1780548095703125,
+      "learning_rate": 0.0001899722104469879,
+      "loss": 2.2257,
+      "step": 2130
+    },
+    {
+      "epoch": 0.05039800292025811,
+      "grad_norm": 2.254702091217041,
+      "learning_rate": 0.00018992510950967926,
+      "loss": 2.175,
+      "step": 2140
+    },
+    {
+      "epoch": 0.050633507606801374,
+      "grad_norm": 2.233217477798462,
+      "learning_rate": 0.00018987800857237058,
+      "loss": 2.2972,
+      "step": 2150
+    },
+    {
+      "epoch": 0.05086901229334464,
+      "grad_norm": 2.3001439571380615,
+      "learning_rate": 0.00018983090763506196,
+      "loss": 2.3244,
+      "step": 2160
+    },
+    {
+      "epoch": 0.0511045169798879,
+      "grad_norm": 2.1131844520568848,
+      "learning_rate": 0.00018978851679148415,
+      "loss": 2.4339,
+      "step": 2170
+    },
+    {
+      "epoch": 0.05134002166643116,
+      "grad_norm": 1.8972138166427612,
+      "learning_rate": 0.0001897414158541755,
+      "loss": 2.1409,
+      "step": 2180
+    },
+    {
+      "epoch": 0.051575526352974424,
+      "grad_norm": 2.831122875213623,
+      "learning_rate": 0.00018969431491686685,
+      "loss": 2.4679,
+      "step": 2190
+    },
+    {
+      "epoch": 0.05181103103951769,
+      "grad_norm": 1.9056183099746704,
+      "learning_rate": 0.0001896472139795582,
+      "loss": 2.2394,
+      "step": 2200
+    },
+    {
+      "epoch": 0.052046535726060945,
+      "grad_norm": 2.218238115310669,
+      "learning_rate": 0.00018960011304224955,
+      "loss": 2.0855,
+      "step": 2210
+    },
+    {
+      "epoch": 0.05228204041260421,
+      "grad_norm": 1.6970261335372925,
+      "learning_rate": 0.0001895530121049409,
+      "loss": 2.3911,
+      "step": 2220
+    },
+    {
+      "epoch": 0.052517545099147474,
+      "grad_norm": 1.9589953422546387,
+      "learning_rate": 0.00018950591116763225,
+      "loss": 2.2206,
+      "step": 2230
+    },
+    {
+      "epoch": 0.05275304978569074,
+      "grad_norm": 1.9307024478912354,
+      "learning_rate": 0.0001894588102303236,
+      "loss": 2.3854,
+      "step": 2240
+    },
+    {
+      "epoch": 0.052988554472233995,
+      "grad_norm": 1.5521360635757446,
+      "learning_rate": 0.00018941170929301495,
+      "loss": 2.3013,
+      "step": 2250
+    },
+    {
+      "epoch": 0.05322405915877726,
+      "grad_norm": 2.1857752799987793,
+      "learning_rate": 0.00018936460835570628,
+      "loss": 2.3844,
+      "step": 2260
+    },
+    {
+      "epoch": 0.053459563845320524,
+      "grad_norm": 2.2541661262512207,
+      "learning_rate": 0.00018931750741839765,
+      "loss": 2.1811,
+      "step": 2270
+    },
+    {
+      "epoch": 0.05369506853186378,
+      "grad_norm": 2.868990421295166,
+      "learning_rate": 0.00018927040648108898,
+      "loss": 2.391,
+      "step": 2280
+    },
+    {
+      "epoch": 0.053930573218407045,
+      "grad_norm": 1.7669481039047241,
+      "learning_rate": 0.00018922330554378033,
+      "loss": 2.2318,
+      "step": 2290
+    },
+    {
+      "epoch": 0.05416607790495031,
+      "grad_norm": 2.0685200691223145,
+      "learning_rate": 0.00018917620460647168,
+      "loss": 2.2587,
+      "step": 2300
+    },
+    {
+      "epoch": 0.05440158259149357,
+      "grad_norm": 2.7001943588256836,
+      "learning_rate": 0.00018912910366916303,
+      "loss": 2.2861,
+      "step": 2310
+    },
+    {
+      "epoch": 0.05463708727803683,
+      "grad_norm": 1.7368396520614624,
+      "learning_rate": 0.00018908200273185438,
+      "loss": 2.119,
+      "step": 2320
+    },
+    {
+      "epoch": 0.054872591964580095,
+      "grad_norm": 2.3278701305389404,
+      "learning_rate": 0.00018903490179454573,
+      "loss": 2.1874,
+      "step": 2330
+    },
+    {
+      "epoch": 0.05510809665112336,
+      "grad_norm": 2.0269453525543213,
+      "learning_rate": 0.00018898780085723705,
+      "loss": 2.3772,
+      "step": 2340
+    },
+    {
+      "epoch": 0.055343601337666616,
+      "grad_norm": 1.5783114433288574,
+      "learning_rate": 0.00018894069991992843,
+      "loss": 2.1799,
+      "step": 2350
+    },
+    {
+      "epoch": 0.05557910602420988,
+      "grad_norm": 2.1277246475219727,
+      "learning_rate": 0.00018889359898261975,
+      "loss": 2.1011,
+      "step": 2360
+    },
+    {
+      "epoch": 0.055814610710753144,
+      "grad_norm": 1.865946650505066,
+      "learning_rate": 0.0001888464980453111,
+      "loss": 2.2497,
+      "step": 2370
+    },
+    {
+      "epoch": 0.05605011539729641,
+      "grad_norm": 2.0735392570495605,
+      "learning_rate": 0.00018879939710800245,
+      "loss": 2.2633,
+      "step": 2380
+    },
+    {
+      "epoch": 0.056285620083839666,
+      "grad_norm": 1.9640772342681885,
+      "learning_rate": 0.0001887522961706938,
+      "loss": 2.2787,
+      "step": 2390
+    },
+    {
+      "epoch": 0.05652112477038293,
+      "grad_norm": 1.4566375017166138,
+      "learning_rate": 0.00018870519523338515,
+      "loss": 2.1823,
+      "step": 2400
+    },
+    {
+      "epoch": 0.056756629456926194,
+      "grad_norm": 1.715774416923523,
+      "learning_rate": 0.0001886580942960765,
+      "loss": 2.1143,
+      "step": 2410
+    },
+    {
+      "epoch": 0.05699213414346946,
+      "grad_norm": 4.055193901062012,
+      "learning_rate": 0.00018861099335876785,
+      "loss": 2.4697,
+      "step": 2420
+    },
+    {
+      "epoch": 0.057227638830012716,
+      "grad_norm": 2.1839892864227295,
+      "learning_rate": 0.0001885638924214592,
+      "loss": 2.0825,
+      "step": 2430
+    },
+    {
+      "epoch": 0.05746314351655598,
+      "grad_norm": 1.8288459777832031,
+      "learning_rate": 0.00018851679148415053,
+      "loss": 2.1559,
+      "step": 2440
+    },
+    {
+      "epoch": 0.057698648203099244,
+      "grad_norm": 1.7593369483947754,
+      "learning_rate": 0.0001884696905468419,
+      "loss": 2.1335,
+      "step": 2450
+    },
+    {
+      "epoch": 0.0579341528896425,
+      "grad_norm": 1.703242540359497,
+      "learning_rate": 0.00018842258960953323,
+      "loss": 2.2544,
+      "step": 2460
+    },
+    {
+      "epoch": 0.058169657576185765,
+      "grad_norm": 2.0734493732452393,
+      "learning_rate": 0.00018837548867222458,
+      "loss": 2.2457,
+      "step": 2470
+    },
+    {
+      "epoch": 0.05840516226272903,
+      "grad_norm": 1.746780514717102,
+      "learning_rate": 0.00018832838773491593,
+      "loss": 2.3449,
+      "step": 2480
+    },
+    {
+      "epoch": 0.058640666949272294,
+      "grad_norm": 2.1612048149108887,
+      "learning_rate": 0.00018828128679760728,
+      "loss": 2.37,
+      "step": 2490
+    },
+    {
+      "epoch": 0.05887617163581555,
+      "grad_norm": 1.9451191425323486,
+      "learning_rate": 0.00018823418586029863,
+      "loss": 2.0534,
+      "step": 2500
+    },
+    {
+      "epoch": 0.059111676322358815,
+      "grad_norm": 1.6598166227340698,
+      "learning_rate": 0.00018818708492298998,
+      "loss": 2.2115,
+      "step": 2510
+    },
+    {
+      "epoch": 0.05934718100890208,
+      "grad_norm": 1.8540189266204834,
+      "learning_rate": 0.00018813998398568133,
+      "loss": 2.2075,
+      "step": 2520
+    },
+    {
+      "epoch": 0.05958268569544534,
+      "grad_norm": 1.914438009262085,
+      "learning_rate": 0.00018809288304837268,
+      "loss": 2.4852,
+      "step": 2530
+    },
+    {
+      "epoch": 0.0598181903819886,
+      "grad_norm": 1.941332459449768,
+      "learning_rate": 0.00018804578211106403,
+      "loss": 2.2285,
+      "step": 2540
+    },
+    {
+      "epoch": 0.060053695068531865,
+      "grad_norm": 1.4906508922576904,
+      "learning_rate": 0.00018799868117375535,
+      "loss": 2.2549,
+      "step": 2550
+    },
+    {
+      "epoch": 0.06028919975507513,
+      "grad_norm": 2.5642271041870117,
+      "learning_rate": 0.00018795158023644673,
+      "loss": 2.3277,
+      "step": 2560
+    },
+    {
+      "epoch": 0.060524704441618386,
+      "grad_norm": 2.510061264038086,
+      "learning_rate": 0.00018790447929913805,
+      "loss": 2.2822,
+      "step": 2570
+    },
+    {
+      "epoch": 0.06076020912816165,
+      "grad_norm": 1.7272487878799438,
+      "learning_rate": 0.00018785737836182943,
+      "loss": 2.1391,
+      "step": 2580
+    },
+    {
+      "epoch": 0.060995713814704915,
+      "grad_norm": 1.9994730949401855,
+      "learning_rate": 0.00018781027742452075,
+      "loss": 2.2,
+      "step": 2590
+    },
+    {
+      "epoch": 0.06123121850124817,
+      "grad_norm": 1.5051331520080566,
+      "learning_rate": 0.0001877631764872121,
+      "loss": 2.3043,
+      "step": 2600
+    },
+    {
+      "epoch": 0.061466723187791436,
+      "grad_norm": 1.6909672021865845,
+      "learning_rate": 0.00018771607554990345,
+      "loss": 2.0515,
+      "step": 2610
+    },
+    {
+      "epoch": 0.0617022278743347,
+      "grad_norm": 1.6561975479125977,
+      "learning_rate": 0.0001876689746125948,
+      "loss": 2.3871,
+      "step": 2620
+    },
+    {
+      "epoch": 0.061937732560877964,
+      "grad_norm": 1.6686687469482422,
+      "learning_rate": 0.00018762187367528615,
+      "loss": 2.2119,
+      "step": 2630
+    },
+    {
+      "epoch": 0.06217323724742122,
+      "grad_norm": 1.7708022594451904,
+      "learning_rate": 0.0001875747727379775,
+      "loss": 2.3067,
+      "step": 2640
+    },
+    {
+      "epoch": 0.062408741933964486,
+      "grad_norm": 1.873261570930481,
+      "learning_rate": 0.00018752767180066883,
+      "loss": 2.3496,
+      "step": 2650
+    },
+    {
+      "epoch": 0.06264424662050774,
+      "grad_norm": 1.7290661334991455,
+      "learning_rate": 0.0001874805708633602,
+      "loss": 2.2128,
+      "step": 2660
+    },
+    {
+      "epoch": 0.06287975130705101,
+      "grad_norm": 2.2264304161071777,
+      "learning_rate": 0.00018743346992605153,
+      "loss": 2.1921,
+      "step": 2670
+    },
+    {
+      "epoch": 0.06311525599359427,
+      "grad_norm": 1.793036699295044,
+      "learning_rate": 0.00018738636898874288,
+      "loss": 2.352,
+      "step": 2680
+    },
+    {
+      "epoch": 0.06335076068013754,
+      "grad_norm": 2.034780502319336,
+      "learning_rate": 0.00018733926805143423,
+      "loss": 2.3553,
+      "step": 2690
+    },
+    {
+      "epoch": 0.0635862653666808,
+      "grad_norm": 1.685520887374878,
+      "learning_rate": 0.00018729216711412558,
+      "loss": 2.1674,
+      "step": 2700
+    },
+    {
+      "epoch": 0.06382177005322406,
+      "grad_norm": 2.387843370437622,
+      "learning_rate": 0.00018724506617681693,
+      "loss": 2.2459,
+      "step": 2710
+    },
+    {
+      "epoch": 0.06405727473976733,
+      "grad_norm": 2.0628840923309326,
+      "learning_rate": 0.00018719796523950828,
+      "loss": 2.3385,
+      "step": 2720
+    },
+    {
+      "epoch": 0.06429277942631058,
+      "grad_norm": 1.780592679977417,
+      "learning_rate": 0.0001871508643021996,
+      "loss": 2.0707,
+      "step": 2730
+    },
+    {
+      "epoch": 0.06452828411285384,
+      "grad_norm": 1.9920777082443237,
+      "learning_rate": 0.00018710376336489098,
+      "loss": 2.175,
+      "step": 2740
+    },
+    {
+      "epoch": 0.0647637887993971,
+      "grad_norm": 1.7009214162826538,
+      "learning_rate": 0.0001870566624275823,
+      "loss": 2.0857,
+      "step": 2750
+    },
+    {
+      "epoch": 0.06499929348594037,
+      "grad_norm": 2.2230193614959717,
+      "learning_rate": 0.00018700956149027365,
+      "loss": 2.4301,
+      "step": 2760
+    },
+    {
+      "epoch": 0.06523479817248364,
+      "grad_norm": 1.7082926034927368,
+      "learning_rate": 0.000186962460552965,
+      "loss": 2.2562,
+      "step": 2770
+    },
+    {
+      "epoch": 0.0654703028590269,
+      "grad_norm": 1.9396960735321045,
+      "learning_rate": 0.00018691535961565635,
+      "loss": 2.3725,
+      "step": 2780
+    },
+    {
+      "epoch": 0.06570580754557016,
+      "grad_norm": 1.9210023880004883,
+      "learning_rate": 0.00018686825867834773,
+      "loss": 2.197,
+      "step": 2790
+    },
+    {
+      "epoch": 0.06594131223211341,
+      "grad_norm": 1.7751202583312988,
+      "learning_rate": 0.00018682115774103905,
+      "loss": 2.4285,
+      "step": 2800
+    },
+    {
+      "epoch": 0.06617681691865668,
+      "grad_norm": 2.46057391166687,
+      "learning_rate": 0.0001867740568037304,
+      "loss": 2.0906,
+      "step": 2810
+    },
+    {
+      "epoch": 0.06641232160519994,
+      "grad_norm": 1.982324481010437,
+      "learning_rate": 0.00018672695586642175,
+      "loss": 2.3326,
+      "step": 2820
+    },
+    {
+      "epoch": 0.0666478262917432,
+      "grad_norm": 2.0937414169311523,
+      "learning_rate": 0.0001866798549291131,
+      "loss": 2.1654,
+      "step": 2830
+    },
+    {
+      "epoch": 0.06688333097828647,
+      "grad_norm": 1.8799653053283691,
+      "learning_rate": 0.00018663275399180446,
+      "loss": 2.178,
+      "step": 2840
+    },
+    {
+      "epoch": 0.06711883566482973,
+      "grad_norm": 2.4298360347747803,
+      "learning_rate": 0.0001865856530544958,
+      "loss": 2.3419,
+      "step": 2850
+    },
+    {
+      "epoch": 0.067354340351373,
+      "grad_norm": 1.9179294109344482,
+      "learning_rate": 0.00018653855211718713,
+      "loss": 2.3138,
+      "step": 2860
+    },
+    {
+      "epoch": 0.06758984503791625,
+      "grad_norm": 1.727042317390442,
+      "learning_rate": 0.0001864914511798785,
+      "loss": 2.1584,
+      "step": 2870
+    },
+    {
+      "epoch": 0.06782534972445951,
+      "grad_norm": 5.538887023925781,
+      "learning_rate": 0.00018644435024256983,
+      "loss": 2.1706,
+      "step": 2880
+    },
+    {
+      "epoch": 0.06806085441100278,
+      "grad_norm": 1.6711513996124268,
+      "learning_rate": 0.00018639724930526118,
+      "loss": 2.1269,
+      "step": 2890
+    },
+    {
+      "epoch": 0.06829635909754604,
+      "grad_norm": 3.3184592723846436,
+      "learning_rate": 0.00018635014836795253,
+      "loss": 2.201,
+      "step": 2900
+    },
+    {
+      "epoch": 0.0685318637840893,
+      "grad_norm": 2.0365543365478516,
+      "learning_rate": 0.00018630304743064388,
+      "loss": 2.203,
+      "step": 2910
+    },
+    {
+      "epoch": 0.06876736847063257,
+      "grad_norm": 2.007431745529175,
+      "learning_rate": 0.00018625594649333523,
+      "loss": 2.1961,
+      "step": 2920
+    },
+    {
+      "epoch": 0.06900287315717583,
+      "grad_norm": 5.275245189666748,
+      "learning_rate": 0.00018620884555602658,
+      "loss": 1.9558,
+      "step": 2930
+    },
+    {
+      "epoch": 0.06923837784371908,
+      "grad_norm": 1.7514394521713257,
+      "learning_rate": 0.0001861617446187179,
+      "loss": 2.2311,
+      "step": 2940
+    },
+    {
+      "epoch": 0.06947388253026235,
+      "grad_norm": 1.685663104057312,
+      "learning_rate": 0.00018611464368140928,
+      "loss": 2.4731,
+      "step": 2950
+    },
+    {
+      "epoch": 0.06970938721680561,
+      "grad_norm": 1.8519153594970703,
+      "learning_rate": 0.0001860675427441006,
+      "loss": 2.3555,
+      "step": 2960
+    },
+    {
+      "epoch": 0.06994489190334888,
+      "grad_norm": 1.6579869985580444,
+      "learning_rate": 0.00018602515190052282,
+      "loss": 2.1312,
+      "step": 2970
+    },
+    {
+      "epoch": 0.07018039658989214,
+      "grad_norm": 1.7049615383148193,
+      "learning_rate": 0.00018597805096321417,
+      "loss": 2.1154,
+      "step": 2980
+    },
+    {
+      "epoch": 0.0704159012764354,
+      "grad_norm": 2.0084738731384277,
+      "learning_rate": 0.00018593095002590552,
+      "loss": 2.1103,
+      "step": 2990
+    },
+    {
+      "epoch": 0.07065140596297867,
+      "grad_norm": 2.9388327598571777,
+      "learning_rate": 0.00018588384908859687,
+      "loss": 2.2574,
+      "step": 3000
+    },
+    {
+      "epoch": 0.07088691064952192,
+      "grad_norm": 1.6756362915039062,
+      "learning_rate": 0.00018583674815128822,
+      "loss": 2.1279,
+      "step": 3010
+    },
+    {
+      "epoch": 0.07112241533606518,
+      "grad_norm": 2.137333393096924,
+      "learning_rate": 0.00018578964721397955,
+      "loss": 2.3244,
+      "step": 3020
+    },
+    {
+      "epoch": 0.07135792002260845,
+      "grad_norm": 1.6805435419082642,
+      "learning_rate": 0.00018574254627667092,
+      "loss": 2.3783,
+      "step": 3030
+    },
+    {
+      "epoch": 0.07159342470915171,
+      "grad_norm": 1.4257123470306396,
+      "learning_rate": 0.00018569544533936227,
+      "loss": 2.1998,
+      "step": 3040
+    },
+    {
+      "epoch": 0.07182892939569498,
+      "grad_norm": 2.6023623943328857,
+      "learning_rate": 0.0001856483444020536,
+      "loss": 2.2162,
+      "step": 3050
+    },
+    {
+      "epoch": 0.07206443408223824,
+      "grad_norm": 1.765554666519165,
+      "learning_rate": 0.00018560124346474497,
+      "loss": 2.1655,
+      "step": 3060
+    },
+    {
+      "epoch": 0.0722999387687815,
+      "grad_norm": 1.852608561515808,
+      "learning_rate": 0.0001855541425274363,
+      "loss": 2.1649,
+      "step": 3070
+    },
+    {
+      "epoch": 0.07253544345532476,
+      "grad_norm": 1.706581950187683,
+      "learning_rate": 0.00018550704159012767,
+      "loss": 2.1517,
+      "step": 3080
+    },
+    {
+      "epoch": 0.07277094814186802,
+      "grad_norm": 2.3597655296325684,
+      "learning_rate": 0.000185459940652819,
+      "loss": 2.0139,
+      "step": 3090
+    },
+    {
+      "epoch": 0.07300645282841128,
+      "grad_norm": 2.2678277492523193,
+      "learning_rate": 0.00018541283971551035,
+      "loss": 2.0053,
+      "step": 3100
+    },
+    {
+      "epoch": 0.07324195751495455,
+      "grad_norm": 2.10081148147583,
+      "learning_rate": 0.0001853657387782017,
+      "loss": 2.2485,
+      "step": 3110
+    },
+    {
+      "epoch": 0.07347746220149781,
+      "grad_norm": 1.7612814903259277,
+      "learning_rate": 0.00018531863784089305,
+      "loss": 2.2997,
+      "step": 3120
+    },
+    {
+      "epoch": 0.07371296688804108,
+      "grad_norm": 1.8904122114181519,
+      "learning_rate": 0.0001852715369035844,
+      "loss": 2.1279,
+      "step": 3130
+    },
+    {
+      "epoch": 0.07394847157458434,
+      "grad_norm": 1.769583821296692,
+      "learning_rate": 0.00018522443596627575,
+      "loss": 2.2626,
+      "step": 3140
+    },
+    {
+      "epoch": 0.07418397626112759,
+      "grad_norm": 2.3224611282348633,
+      "learning_rate": 0.00018517733502896707,
+      "loss": 2.3991,
+      "step": 3150
+    },
+    {
+      "epoch": 0.07441948094767085,
+      "grad_norm": 1.9851993322372437,
+      "learning_rate": 0.00018513023409165845,
+      "loss": 2.1701,
+      "step": 3160
+    },
+    {
+      "epoch": 0.07465498563421412,
+      "grad_norm": 2.010765790939331,
+      "learning_rate": 0.00018508313315434977,
+      "loss": 2.2649,
+      "step": 3170
+    },
+    {
+      "epoch": 0.07489049032075738,
+      "grad_norm": 1.9443213939666748,
+      "learning_rate": 0.00018503603221704112,
+      "loss": 2.2097,
+      "step": 3180
+    },
+    {
+      "epoch": 0.07512599500730065,
+      "grad_norm": 2.350292682647705,
+      "learning_rate": 0.00018498893127973247,
+      "loss": 2.2244,
+      "step": 3190
+    },
+    {
+      "epoch": 0.07536149969384391,
+      "grad_norm": 1.83067786693573,
+      "learning_rate": 0.00018494183034242382,
+      "loss": 2.0924,
+      "step": 3200
+    },
+    {
+      "epoch": 0.07559700438038718,
+      "grad_norm": 2.046450614929199,
+      "learning_rate": 0.00018489472940511517,
+      "loss": 2.214,
+      "step": 3210
+    },
+    {
+      "epoch": 0.07583250906693043,
+      "grad_norm": 2.111867666244507,
+      "learning_rate": 0.00018484762846780652,
+      "loss": 2.5115,
+      "step": 3220
+    },
+    {
+      "epoch": 0.07606801375347369,
+      "grad_norm": 1.992246150970459,
+      "learning_rate": 0.00018480052753049785,
+      "loss": 2.1997,
+      "step": 3230
+    },
+    {
+      "epoch": 0.07630351844001695,
+      "grad_norm": 2.763298511505127,
+      "learning_rate": 0.00018475342659318922,
+      "loss": 2.299,
+      "step": 3240
+    },
+    {
+      "epoch": 0.07653902312656022,
+      "grad_norm": 1.8757209777832031,
+      "learning_rate": 0.00018470632565588055,
+      "loss": 2.323,
+      "step": 3250
+    },
+    {
+      "epoch": 0.07677452781310348,
+      "grad_norm": 1.779782772064209,
+      "learning_rate": 0.0001846592247185719,
+      "loss": 2.0727,
+      "step": 3260
+    },
+    {
+      "epoch": 0.07701003249964675,
+      "grad_norm": 2.201812267303467,
+      "learning_rate": 0.00018461212378126325,
+      "loss": 2.4871,
+      "step": 3270
+    },
+    {
+      "epoch": 0.07724553718619001,
+      "grad_norm": 2.0720343589782715,
+      "learning_rate": 0.0001845650228439546,
+      "loss": 2.1938,
+      "step": 3280
+    },
+    {
+      "epoch": 0.07748104187273326,
+      "grad_norm": 1.743800163269043,
+      "learning_rate": 0.00018451792190664595,
+      "loss": 2.2414,
+      "step": 3290
+    },
+    {
+      "epoch": 0.07771654655927653,
+      "grad_norm": 1.9348535537719727,
+      "learning_rate": 0.0001844708209693373,
+      "loss": 2.0799,
+      "step": 3300
+    },
+    {
+      "epoch": 0.07795205124581979,
+      "grad_norm": 2.4140963554382324,
+      "learning_rate": 0.00018442372003202862,
+      "loss": 2.1721,
+      "step": 3310
+    },
+    {
+      "epoch": 0.07818755593236305,
+      "grad_norm": 1.6908040046691895,
+      "learning_rate": 0.00018437661909472,
+      "loss": 2.0753,
+      "step": 3320
+    },
+    {
+      "epoch": 0.07842306061890632,
+      "grad_norm": 2.733171224594116,
+      "learning_rate": 0.00018432951815741135,
+      "loss": 2.1139,
+      "step": 3330
+    },
+    {
+      "epoch": 0.07865856530544958,
+      "grad_norm": 2.480011463165283,
+      "learning_rate": 0.0001842824172201027,
+      "loss": 2.2136,
+      "step": 3340
+    },
+    {
+      "epoch": 0.07889406999199285,
+      "grad_norm": 2.223186731338501,
+      "learning_rate": 0.00018423531628279405,
+      "loss": 1.8241,
+      "step": 3350
+    },
+    {
+      "epoch": 0.0791295746785361,
+      "grad_norm": 2.359806537628174,
+      "learning_rate": 0.00018418821534548537,
+      "loss": 2.2171,
+      "step": 3360
+    },
+    {
+      "epoch": 0.07936507936507936,
+      "grad_norm": 1.6184040307998657,
+      "learning_rate": 0.00018414111440817675,
+      "loss": 2.3467,
+      "step": 3370
+    },
+    {
+      "epoch": 0.07960058405162262,
+      "grad_norm": 1.9266674518585205,
+      "learning_rate": 0.00018409401347086807,
+      "loss": 2.092,
+      "step": 3380
+    },
+    {
+      "epoch": 0.07983608873816589,
+      "grad_norm": 3.338719129562378,
+      "learning_rate": 0.00018404691253355942,
+      "loss": 2.2806,
+      "step": 3390
+    },
+    {
+      "epoch": 0.08007159342470915,
+      "grad_norm": 2.559915781021118,
+      "learning_rate": 0.00018399981159625077,
+      "loss": 2.346,
+      "step": 3400
+    },
+    {
+      "epoch": 0.08030709811125242,
+      "grad_norm": 2.17266845703125,
+      "learning_rate": 0.00018395271065894212,
+      "loss": 2.3376,
+      "step": 3410
+    },
+    {
+      "epoch": 0.08054260279779568,
+      "grad_norm": 1.8615611791610718,
+      "learning_rate": 0.00018390560972163347,
+      "loss": 2.204,
+      "step": 3420
+    },
+    {
+      "epoch": 0.08077810748433893,
+      "grad_norm": 1.8173818588256836,
+      "learning_rate": 0.00018385850878432482,
+      "loss": 2.1612,
+      "step": 3430
+    },
+    {
+      "epoch": 0.0810136121708822,
+      "grad_norm": 2.4427809715270996,
+      "learning_rate": 0.00018381140784701615,
+      "loss": 2.379,
+      "step": 3440
+    },
+    {
+      "epoch": 0.08124911685742546,
+      "grad_norm": 1.7036974430084229,
+      "learning_rate": 0.00018376430690970753,
+      "loss": 2.3024,
+      "step": 3450
+    },
+    {
+      "epoch": 0.08148462154396872,
+      "grad_norm": 2.6452062129974365,
+      "learning_rate": 0.00018371720597239885,
+      "loss": 2.0425,
+      "step": 3460
+    },
+    {
+      "epoch": 0.08172012623051199,
+      "grad_norm": 1.7686418294906616,
+      "learning_rate": 0.0001836701050350902,
+      "loss": 2.1656,
+      "step": 3470
+    },
+    {
+      "epoch": 0.08195563091705525,
+      "grad_norm": 1.9566192626953125,
+      "learning_rate": 0.00018362300409778155,
+      "loss": 2.3603,
+      "step": 3480
+    },
+    {
+      "epoch": 0.08219113560359852,
+      "grad_norm": 1.8868483304977417,
+      "learning_rate": 0.0001835759031604729,
+      "loss": 2.2435,
+      "step": 3490
+    },
+    {
+      "epoch": 0.08242664029014177,
+      "grad_norm": 2.2280383110046387,
+      "learning_rate": 0.00018352880222316425,
+      "loss": 2.2404,
+      "step": 3500
+    },
+    {
+      "epoch": 0.08266214497668503,
+      "grad_norm": 2.1244733333587646,
+      "learning_rate": 0.0001834817012858556,
+      "loss": 2.0227,
+      "step": 3510
+    },
+    {
+      "epoch": 0.0828976496632283,
+      "grad_norm": 1.9479572772979736,
+      "learning_rate": 0.00018343460034854695,
+      "loss": 2.0774,
+      "step": 3520
+    },
+    {
+      "epoch": 0.08313315434977156,
+      "grad_norm": 1.8838822841644287,
+      "learning_rate": 0.0001833874994112383,
+      "loss": 2.3049,
+      "step": 3530
+    },
+    {
+      "epoch": 0.08336865903631482,
+      "grad_norm": 2.044419288635254,
+      "learning_rate": 0.00018334039847392962,
+      "loss": 2.0416,
+      "step": 3540
+    },
+    {
+      "epoch": 0.08360416372285809,
+      "grad_norm": 2.2245490550994873,
+      "learning_rate": 0.000183293297536621,
+      "loss": 2.13,
+      "step": 3550
+    },
+    {
+      "epoch": 0.08383966840940135,
+      "grad_norm": 2.318081855773926,
+      "learning_rate": 0.00018324619659931232,
+      "loss": 2.1373,
+      "step": 3560
+    },
+    {
+      "epoch": 0.0840751730959446,
+      "grad_norm": 1.8568158149719238,
+      "learning_rate": 0.00018319909566200367,
+      "loss": 2.2176,
+      "step": 3570
+    },
+    {
+      "epoch": 0.08431067778248787,
+      "grad_norm": 2.3757684230804443,
+      "learning_rate": 0.00018315199472469502,
+      "loss": 2.299,
+      "step": 3580
+    },
+    {
+      "epoch": 0.08454618246903113,
+      "grad_norm": 3.102747917175293,
+      "learning_rate": 0.00018310489378738637,
+      "loss": 2.1751,
+      "step": 3590
+    },
+    {
+      "epoch": 0.0847816871555744,
+      "grad_norm": 2.0704846382141113,
+      "learning_rate": 0.00018305779285007772,
+      "loss": 2.2471,
+      "step": 3600
+    },
+    {
+      "epoch": 0.08501719184211766,
+      "grad_norm": 1.9258019924163818,
+      "learning_rate": 0.00018301069191276908,
+      "loss": 2.1349,
+      "step": 3610
+    },
+    {
+      "epoch": 0.08525269652866092,
+      "grad_norm": 2.328375816345215,
+      "learning_rate": 0.00018296359097546043,
+      "loss": 2.3323,
+      "step": 3620
+    },
+    {
+      "epoch": 0.08548820121520419,
+      "grad_norm": 1.6533223390579224,
+      "learning_rate": 0.00018291649003815178,
+      "loss": 2.3557,
+      "step": 3630
+    },
+    {
+      "epoch": 0.08572370590174744,
+      "grad_norm": 1.7346618175506592,
+      "learning_rate": 0.00018286938910084313,
+      "loss": 2.1023,
+      "step": 3640
+    },
+    {
+      "epoch": 0.0859592105882907,
+      "grad_norm": 2.2735652923583984,
+      "learning_rate": 0.00018282228816353445,
+      "loss": 2.2031,
+      "step": 3650
+    },
+    {
+      "epoch": 0.08619471527483397,
+      "grad_norm": 1.8555744886398315,
+      "learning_rate": 0.00018277518722622583,
+      "loss": 2.2743,
+      "step": 3660
+    },
+    {
+      "epoch": 0.08643021996137723,
+      "grad_norm": 1.919329285621643,
+      "learning_rate": 0.00018272808628891715,
+      "loss": 2.2808,
+      "step": 3670
+    },
+    {
+      "epoch": 0.0866657246479205,
+      "grad_norm": 2.246380090713501,
+      "learning_rate": 0.00018268098535160853,
+      "loss": 2.3911,
+      "step": 3680
+    },
+    {
+      "epoch": 0.08690122933446376,
+      "grad_norm": 1.6249585151672363,
+      "learning_rate": 0.00018263388441429985,
+      "loss": 2.0467,
+      "step": 3690
+    },
+    {
+      "epoch": 0.08713673402100702,
+      "grad_norm": 1.8340191841125488,
+      "learning_rate": 0.0001825867834769912,
+      "loss": 2.188,
+      "step": 3700
+    },
+    {
+      "epoch": 0.08737223870755029,
+      "grad_norm": 2.3175435066223145,
+      "learning_rate": 0.00018253968253968255,
+      "loss": 2.2183,
+      "step": 3710
+    },
+    {
+      "epoch": 0.08760774339409354,
+      "grad_norm": 2.404165267944336,
+      "learning_rate": 0.0001824925816023739,
+      "loss": 2.0468,
+      "step": 3720
+    },
+    {
+      "epoch": 0.0878432480806368,
+      "grad_norm": 2.1523420810699463,
+      "learning_rate": 0.00018244548066506525,
+      "loss": 2.1458,
+      "step": 3730
+    },
+    {
+      "epoch": 0.08807875276718007,
+      "grad_norm": 2.125816583633423,
+      "learning_rate": 0.0001823983797277566,
+      "loss": 2.3276,
+      "step": 3740
+    },
+    {
+      "epoch": 0.08831425745372333,
+      "grad_norm": 2.124753713607788,
+      "learning_rate": 0.00018235127879044792,
+      "loss": 2.222,
+      "step": 3750
+    },
+    {
+      "epoch": 0.0885497621402666,
+      "grad_norm": 2.3880615234375,
+      "learning_rate": 0.0001823041778531393,
+      "loss": 2.1973,
+      "step": 3760
+    },
+    {
+      "epoch": 0.08878526682680986,
+      "grad_norm": 1.9938665628433228,
+      "learning_rate": 0.00018225707691583062,
+      "loss": 2.0841,
+      "step": 3770
+    },
+    {
+      "epoch": 0.08902077151335312,
+      "grad_norm": 2.1049611568450928,
+      "learning_rate": 0.00018220997597852197,
+      "loss": 2.364,
+      "step": 3780
+    },
+    {
+      "epoch": 0.08925627619989637,
+      "grad_norm": 2.110180139541626,
+      "learning_rate": 0.00018216287504121333,
+      "loss": 2.346,
+      "step": 3790
+    },
+    {
+      "epoch": 0.08949178088643964,
+      "grad_norm": 1.598199725151062,
+      "learning_rate": 0.00018211577410390468,
+      "loss": 2.0503,
+      "step": 3800
+    },
+    {
+      "epoch": 0.0897272855729829,
+      "grad_norm": 2.412278175354004,
+      "learning_rate": 0.00018206867316659603,
+      "loss": 2.2061,
+      "step": 3810
+    },
+    {
+      "epoch": 0.08996279025952617,
+      "grad_norm": 2.158580780029297,
+      "learning_rate": 0.00018202157222928738,
+      "loss": 2.1454,
+      "step": 3820
+    },
+    {
+      "epoch": 0.09019829494606943,
+      "grad_norm": 1.819800853729248,
+      "learning_rate": 0.0001819744712919787,
+      "loss": 2.4598,
+      "step": 3830
+    },
+    {
+      "epoch": 0.0904337996326127,
+      "grad_norm": 1.7876063585281372,
+      "learning_rate": 0.00018192737035467008,
+      "loss": 2.1973,
+      "step": 3840
+    },
+    {
+      "epoch": 0.09066930431915596,
+      "grad_norm": 2.018787384033203,
+      "learning_rate": 0.0001818802694173614,
+      "loss": 2.209,
+      "step": 3850
+    },
+    {
+      "epoch": 0.09090480900569921,
+      "grad_norm": 1.9427393674850464,
+      "learning_rate": 0.00018183316848005275,
+      "loss": 2.064,
+      "step": 3860
+    },
+    {
+      "epoch": 0.09114031369224247,
+      "grad_norm": 2.2805685997009277,
+      "learning_rate": 0.0001817860675427441,
+      "loss": 2.2278,
+      "step": 3870
+    },
+    {
+      "epoch": 0.09137581837878574,
+      "grad_norm": 1.9191609621047974,
+      "learning_rate": 0.00018173896660543545,
+      "loss": 2.1242,
+      "step": 3880
+    },
+    {
+      "epoch": 0.091611323065329,
+      "grad_norm": 2.0127153396606445,
+      "learning_rate": 0.00018169186566812683,
+      "loss": 2.1634,
+      "step": 3890
+    },
+    {
+      "epoch": 0.09184682775187226,
+      "grad_norm": 1.8756777048110962,
+      "learning_rate": 0.00018164476473081815,
+      "loss": 2.0746,
+      "step": 3900
+    },
+    {
+      "epoch": 0.09208233243841553,
+      "grad_norm": 1.7346595525741577,
+      "learning_rate": 0.0001815976637935095,
+      "loss": 2.1975,
+      "step": 3910
+    },
+    {
+      "epoch": 0.0923178371249588,
+      "grad_norm": 2.1969356536865234,
+      "learning_rate": 0.00018155056285620085,
+      "loss": 2.1953,
+      "step": 3920
+    },
+    {
+      "epoch": 0.09255334181150204,
+      "grad_norm": 2.3179337978363037,
+      "learning_rate": 0.0001815034619188922,
+      "loss": 2.2919,
+      "step": 3930
+    },
+    {
+      "epoch": 0.09278884649804531,
+      "grad_norm": 2.3929030895233154,
+      "learning_rate": 0.00018145636098158355,
+      "loss": 2.184,
+      "step": 3940
+    },
+    {
+      "epoch": 0.09302435118458857,
+      "grad_norm": 2.515260934829712,
+      "learning_rate": 0.0001814092600442749,
+      "loss": 2.2885,
+      "step": 3950
+    },
+    {
+      "epoch": 0.09325985587113184,
+      "grad_norm": 2.0467185974121094,
+      "learning_rate": 0.00018136215910696623,
+      "loss": 2.2213,
+      "step": 3960
+    },
+    {
+      "epoch": 0.0934953605576751,
+      "grad_norm": 1.7174040079116821,
+      "learning_rate": 0.0001813150581696576,
+      "loss": 2.0542,
+      "step": 3970
+    },
+    {
+      "epoch": 0.09373086524421836,
+      "grad_norm": 2.1928911209106445,
+      "learning_rate": 0.00018126795723234893,
+      "loss": 2.0091,
+      "step": 3980
+    },
+    {
+      "epoch": 0.09396636993076163,
+      "grad_norm": 2.9600062370300293,
+      "learning_rate": 0.00018122085629504028,
+      "loss": 2.3568,
+      "step": 3990
+    },
+    {
+      "epoch": 0.09420187461730488,
+      "grad_norm": 1.843605637550354,
+      "learning_rate": 0.00018117375535773163,
+      "loss": 2.1123,
+      "step": 4000
+    },
+    {
+      "epoch": 0.09443737930384814,
+      "grad_norm": 1.823076844215393,
+      "learning_rate": 0.00018112665442042298,
+      "loss": 2.5867,
+      "step": 4010
+    },
+    {
+      "epoch": 0.09467288399039141,
+      "grad_norm": 2.1632487773895264,
+      "learning_rate": 0.00018107955348311433,
+      "loss": 2.1294,
+      "step": 4020
+    },
+    {
+      "epoch": 0.09490838867693467,
+      "grad_norm": 1.726159930229187,
+      "learning_rate": 0.00018103245254580568,
+      "loss": 2.2689,
+      "step": 4030
+    },
+    {
+      "epoch": 0.09514389336347794,
+      "grad_norm": 2.210451364517212,
+      "learning_rate": 0.000180985351608497,
+      "loss": 2.044,
+      "step": 4040
+    },
+    {
+      "epoch": 0.0953793980500212,
+      "grad_norm": 2.0674283504486084,
+      "learning_rate": 0.00018093825067118838,
+      "loss": 2.2215,
+      "step": 4050
+    },
+    {
+      "epoch": 0.09561490273656446,
+      "grad_norm": 2.5523428916931152,
+      "learning_rate": 0.0001808911497338797,
+      "loss": 1.9213,
+      "step": 4060
+    },
+    {
+      "epoch": 0.09585040742310771,
+      "grad_norm": 2.207009792327881,
+      "learning_rate": 0.00018084404879657105,
+      "loss": 2.4428,
+      "step": 4070
+    },
+    {
+      "epoch": 0.09608591210965098,
+      "grad_norm": 1.6344585418701172,
+      "learning_rate": 0.0001807969478592624,
+      "loss": 2.0769,
+      "step": 4080
+    },
+    {
+      "epoch": 0.09632141679619424,
+      "grad_norm": 1.8036699295043945,
+      "learning_rate": 0.00018074984692195375,
+      "loss": 2.3486,
+      "step": 4090
+    },
+    {
+      "epoch": 0.0965569214827375,
+      "grad_norm": 1.708190679550171,
+      "learning_rate": 0.0001807027459846451,
+      "loss": 1.9418,
+      "step": 4100
+    },
+    {
+      "epoch": 0.09679242616928077,
+      "grad_norm": 1.9009705781936646,
+      "learning_rate": 0.00018065564504733645,
+      "loss": 2.2658,
+      "step": 4110
+    },
+    {
+      "epoch": 0.09702793085582403,
+      "grad_norm": 1.9692081212997437,
+      "learning_rate": 0.0001806085441100278,
+      "loss": 1.9867,
+      "step": 4120
+    },
+    {
+      "epoch": 0.0972634355423673,
+      "grad_norm": 1.6455093622207642,
+      "learning_rate": 0.00018056144317271915,
+      "loss": 2.2132,
+      "step": 4130
+    },
+    {
+      "epoch": 0.09749894022891055,
+      "grad_norm": 1.9711624383926392,
+      "learning_rate": 0.00018051434223541048,
+      "loss": 2.3335,
+      "step": 4140
+    },
+    {
+      "epoch": 0.09773444491545381,
+      "grad_norm": 2.410085678100586,
+      "learning_rate": 0.00018046724129810185,
+      "loss": 2.1966,
+      "step": 4150
+    },
+    {
+      "epoch": 0.09796994960199708,
+      "grad_norm": 1.9884003400802612,
+      "learning_rate": 0.0001804201403607932,
+      "loss": 2.2239,
+      "step": 4160
+    },
+    {
+      "epoch": 0.09820545428854034,
+      "grad_norm": 1.7518419027328491,
+      "learning_rate": 0.00018037303942348453,
+      "loss": 2.2178,
+      "step": 4170
+    },
+    {
+      "epoch": 0.0984409589750836,
+      "grad_norm": 2.0874931812286377,
+      "learning_rate": 0.0001803259384861759,
+      "loss": 2.0643,
+      "step": 4180
+    },
+    {
+      "epoch": 0.09867646366162687,
+      "grad_norm": 2.409583568572998,
+      "learning_rate": 0.00018027883754886723,
+      "loss": 2.2727,
+      "step": 4190
+    },
+    {
+      "epoch": 0.09891196834817013,
+      "grad_norm": 2.025693655014038,
+      "learning_rate": 0.00018023173661155858,
+      "loss": 2.0157,
+      "step": 4200
+    },
+    {
+      "epoch": 0.09914747303471338,
+      "grad_norm": 1.8936916589736938,
+      "learning_rate": 0.00018018463567424993,
+      "loss": 2.0757,
+      "step": 4210
+    },
+    {
+      "epoch": 0.09938297772125665,
+      "grad_norm": 1.8962174654006958,
+      "learning_rate": 0.00018013753473694128,
+      "loss": 2.1957,
+      "step": 4220
+    },
+    {
+      "epoch": 0.09961848240779991,
+      "grad_norm": 2.055262565612793,
+      "learning_rate": 0.00018009043379963263,
+      "loss": 2.1943,
+      "step": 4230
+    },
+    {
+      "epoch": 0.09985398709434318,
+      "grad_norm": 2.352290153503418,
+      "learning_rate": 0.00018004333286232398,
+      "loss": 2.3158,
+      "step": 4240
+    },
+    {
+      "epoch": 0.10008949178088644,
+      "grad_norm": 1.8247345685958862,
+      "learning_rate": 0.0001799962319250153,
+      "loss": 2.1219,
+      "step": 4250
+    },
+    {
+      "epoch": 0.1003249964674297,
+      "grad_norm": 2.0612590312957764,
+      "learning_rate": 0.00017994913098770668,
+      "loss": 2.1914,
+      "step": 4260
+    },
+    {
+      "epoch": 0.10056050115397297,
+      "grad_norm": 2.3253748416900635,
+      "learning_rate": 0.000179902030050398,
+      "loss": 2.2833,
+      "step": 4270
+    },
+    {
+      "epoch": 0.10079600584051622,
+      "grad_norm": 2.5230417251586914,
+      "learning_rate": 0.00017985492911308938,
+      "loss": 2.2253,
+      "step": 4280
+    },
+    {
+      "epoch": 0.10103151052705948,
+      "grad_norm": 2.022261142730713,
+      "learning_rate": 0.0001798078281757807,
+      "loss": 2.0184,
+      "step": 4290
+    },
+    {
+      "epoch": 0.10126701521360275,
+      "grad_norm": 3.5165741443634033,
+      "learning_rate": 0.00017976072723847205,
+      "loss": 2.2235,
+      "step": 4300
+    },
+    {
+      "epoch": 0.10150251990014601,
+      "grad_norm": 2.481456756591797,
+      "learning_rate": 0.0001797136263011634,
+      "loss": 2.3874,
+      "step": 4310
+    },
+    {
+      "epoch": 0.10173802458668928,
+      "grad_norm": 1.780819058418274,
+      "learning_rate": 0.00017966652536385475,
+      "loss": 2.2076,
+      "step": 4320
+    },
+    {
+      "epoch": 0.10197352927323254,
+      "grad_norm": 1.6852662563323975,
+      "learning_rate": 0.0001796194244265461,
+      "loss": 2.1545,
+      "step": 4330
+    },
+    {
+      "epoch": 0.1022090339597758,
+      "grad_norm": 2.13392972946167,
+      "learning_rate": 0.00017957232348923745,
+      "loss": 2.2223,
+      "step": 4340
+    },
+    {
+      "epoch": 0.10244453864631906,
+      "grad_norm": 1.8714747428894043,
+      "learning_rate": 0.00017952522255192878,
+      "loss": 2.0202,
+      "step": 4350
+    },
+    {
+      "epoch": 0.10268004333286232,
+      "grad_norm": 1.8120633363723755,
+      "learning_rate": 0.00017947812161462015,
+      "loss": 2.2459,
+      "step": 4360
+    },
+    {
+      "epoch": 0.10291554801940558,
+      "grad_norm": 2.3447976112365723,
+      "learning_rate": 0.00017943102067731148,
+      "loss": 2.1867,
+      "step": 4370
+    },
+    {
+      "epoch": 0.10315105270594885,
+      "grad_norm": 2.4707727432250977,
+      "learning_rate": 0.00017938391974000283,
+      "loss": 2.3336,
+      "step": 4380
+    },
+    {
+      "epoch": 0.10338655739249211,
+      "grad_norm": 1.8624794483184814,
+      "learning_rate": 0.00017933681880269418,
+      "loss": 2.2378,
+      "step": 4390
+    },
+    {
+      "epoch": 0.10362206207903538,
+      "grad_norm": 1.687395691871643,
+      "learning_rate": 0.00017928971786538553,
+      "loss": 2.1106,
+      "step": 4400
+    },
+    {
+      "epoch": 0.10385756676557864,
+      "grad_norm": 1.6846327781677246,
+      "learning_rate": 0.00017924261692807688,
+      "loss": 2.0948,
+      "step": 4410
+    },
+    {
+      "epoch": 0.10409307145212189,
+      "grad_norm": 1.7584487199783325,
+      "learning_rate": 0.00017919551599076823,
+      "loss": 2.2761,
+      "step": 4420
+    },
+    {
+      "epoch": 0.10432857613866516,
+      "grad_norm": 2.1012890338897705,
+      "learning_rate": 0.00017914841505345955,
+      "loss": 2.3109,
+      "step": 4430
+    },
+    {
+      "epoch": 0.10456408082520842,
+      "grad_norm": 1.6275670528411865,
+      "learning_rate": 0.00017910131411615093,
+      "loss": 2.1511,
+      "step": 4440
+    },
+    {
+      "epoch": 0.10479958551175168,
+      "grad_norm": 1.744971513748169,
+      "learning_rate": 0.00017905421317884228,
+      "loss": 2.3064,
+      "step": 4450
+    },
+    {
+      "epoch": 0.10503509019829495,
+      "grad_norm": 1.8816249370574951,
+      "learning_rate": 0.0001790071122415336,
+      "loss": 2.4038,
+      "step": 4460
+    },
+    {
+      "epoch": 0.10527059488483821,
+      "grad_norm": 2.0134501457214355,
+      "learning_rate": 0.00017896001130422498,
+      "loss": 2.3979,
+      "step": 4470
+    },
+    {
+      "epoch": 0.10550609957138148,
+      "grad_norm": 2.5849740505218506,
+      "learning_rate": 0.0001789129103669163,
+      "loss": 2.4298,
+      "step": 4480
+    },
+    {
+      "epoch": 0.10574160425792473,
+      "grad_norm": 1.6863932609558105,
+      "learning_rate": 0.00017886580942960768,
+      "loss": 2.0796,
+      "step": 4490
+    },
+    {
+      "epoch": 0.10597710894446799,
+      "grad_norm": 2.486143112182617,
+      "learning_rate": 0.000178818708492299,
+      "loss": 2.1871,
+      "step": 4500
+    },
+    {
+      "epoch": 0.10621261363101125,
+      "grad_norm": 1.7740905284881592,
+      "learning_rate": 0.00017877160755499035,
+      "loss": 1.9941,
+      "step": 4510
+    },
+    {
+      "epoch": 0.10644811831755452,
+      "grad_norm": 2.0685887336730957,
+      "learning_rate": 0.0001787245066176817,
+      "loss": 2.2425,
+      "step": 4520
+    },
+    {
+      "epoch": 0.10668362300409778,
+      "grad_norm": 1.8682607412338257,
+      "learning_rate": 0.00017867740568037305,
+      "loss": 2.4737,
+      "step": 4530
+    },
+    {
+      "epoch": 0.10691912769064105,
+      "grad_norm": 3.012972116470337,
+      "learning_rate": 0.0001786303047430644,
+      "loss": 2.2102,
+      "step": 4540
+    },
+    {
+      "epoch": 0.10715463237718431,
+      "grad_norm": 2.1660654544830322,
+      "learning_rate": 0.00017858320380575575,
+      "loss": 1.9413,
+      "step": 4550
+    },
+    {
+      "epoch": 0.10739013706372756,
+      "grad_norm": 2.0424931049346924,
+      "learning_rate": 0.00017853610286844708,
+      "loss": 2.1046,
+      "step": 4560
+    },
+    {
+      "epoch": 0.10762564175027083,
+      "grad_norm": 2.7950327396392822,
+      "learning_rate": 0.00017848900193113845,
+      "loss": 2.0865,
+      "step": 4570
+    },
+    {
+      "epoch": 0.10786114643681409,
+      "grad_norm": 1.6264240741729736,
+      "learning_rate": 0.00017844190099382978,
+      "loss": 2.1237,
+      "step": 4580
+    },
+    {
+      "epoch": 0.10809665112335735,
+      "grad_norm": 1.7138210535049438,
+      "learning_rate": 0.00017839480005652113,
+      "loss": 2.2916,
+      "step": 4590
+    },
+    {
+      "epoch": 0.10833215580990062,
+      "grad_norm": 1.89190673828125,
+      "learning_rate": 0.00017834769911921248,
+      "loss": 2.3419,
+      "step": 4600
+    },
+    {
+      "epoch": 0.10856766049644388,
+      "grad_norm": 1.813306212425232,
+      "learning_rate": 0.00017830059818190383,
+      "loss": 2.3194,
+      "step": 4610
+    },
+    {
+      "epoch": 0.10880316518298715,
+      "grad_norm": 1.9797568321228027,
+      "learning_rate": 0.00017825349724459518,
+      "loss": 2.242,
+      "step": 4620
+    },
+    {
+      "epoch": 0.1090386698695304,
+      "grad_norm": 1.423639178276062,
+      "learning_rate": 0.00017820639630728653,
+      "loss": 2.1681,
+      "step": 4630
+    },
+    {
+      "epoch": 0.10927417455607366,
+      "grad_norm": 2.0920000076293945,
+      "learning_rate": 0.00017815929536997785,
+      "loss": 2.155,
+      "step": 4640
+    },
+    {
+      "epoch": 0.10950967924261693,
+      "grad_norm": 2.133775234222412,
+      "learning_rate": 0.00017811219443266923,
+      "loss": 2.1228,
+      "step": 4650
+    },
+    {
+      "epoch": 0.10974518392916019,
+      "grad_norm": 1.449283242225647,
+      "learning_rate": 0.00017806509349536055,
+      "loss": 2.2264,
+      "step": 4660
+    },
+    {
+      "epoch": 0.10998068861570345,
+      "grad_norm": 2.10311222076416,
+      "learning_rate": 0.0001780179925580519,
+      "loss": 2.1021,
+      "step": 4670
+    },
+    {
+      "epoch": 0.11021619330224672,
+      "grad_norm": 1.8965257406234741,
+      "learning_rate": 0.00017797089162074325,
+      "loss": 2.0347,
+      "step": 4680
+    },
+    {
+      "epoch": 0.11045169798878998,
+      "grad_norm": 1.989089846611023,
+      "learning_rate": 0.0001779237906834346,
+      "loss": 2.3597,
+      "step": 4690
+    },
+    {
+      "epoch": 0.11068720267533323,
+      "grad_norm": 2.173595666885376,
+      "learning_rate": 0.00017787668974612595,
+      "loss": 2.3751,
+      "step": 4700
+    },
+    {
+      "epoch": 0.1109227073618765,
+      "grad_norm": 2.043771743774414,
+      "learning_rate": 0.0001778295888088173,
+      "loss": 2.1705,
+      "step": 4710
+    },
+    {
+      "epoch": 0.11115821204841976,
+      "grad_norm": 2.4140965938568115,
+      "learning_rate": 0.00017778248787150865,
+      "loss": 2.2574,
+      "step": 4720
+    },
+    {
+      "epoch": 0.11139371673496302,
+      "grad_norm": 1.8277655839920044,
+      "learning_rate": 0.0001777353869342,
+      "loss": 2.1179,
+      "step": 4730
+    },
+    {
+      "epoch": 0.11162922142150629,
+      "grad_norm": 1.941874623298645,
+      "learning_rate": 0.00017768828599689135,
+      "loss": 2.0465,
+      "step": 4740
+    },
+    {
+      "epoch": 0.11186472610804955,
+      "grad_norm": 1.754234790802002,
+      "learning_rate": 0.0001776411850595827,
+      "loss": 2.2535,
+      "step": 4750
+    },
+    {
+      "epoch": 0.11210023079459282,
+      "grad_norm": 1.6321264505386353,
+      "learning_rate": 0.00017759408412227405,
+      "loss": 2.041,
+      "step": 4760
+    },
+    {
+      "epoch": 0.11233573548113608,
+      "grad_norm": 2.81117582321167,
+      "learning_rate": 0.00017754698318496538,
+      "loss": 2.2673,
+      "step": 4770
+    },
+    {
+      "epoch": 0.11257124016767933,
+      "grad_norm": 1.9449315071105957,
+      "learning_rate": 0.00017749988224765676,
+      "loss": 2.1277,
+      "step": 4780
+    },
+    {
+      "epoch": 0.1128067448542226,
+      "grad_norm": 1.9723138809204102,
+      "learning_rate": 0.00017745278131034808,
+      "loss": 2.141,
+      "step": 4790
+    },
+    {
+      "epoch": 0.11304224954076586,
+      "grad_norm": 2.3215932846069336,
+      "learning_rate": 0.00017740568037303943,
+      "loss": 2.1263,
+      "step": 4800
+    },
+    {
+      "epoch": 0.11327775422730912,
+      "grad_norm": 2.240413188934326,
+      "learning_rate": 0.00017735857943573078,
+      "loss": 2.1786,
+      "step": 4810
+    },
+    {
+      "epoch": 0.11351325891385239,
+      "grad_norm": 2.177074432373047,
+      "learning_rate": 0.00017731147849842213,
+      "loss": 2.1403,
+      "step": 4820
+    },
+    {
+      "epoch": 0.11374876360039565,
+      "grad_norm": 1.7291984558105469,
+      "learning_rate": 0.00017726437756111348,
+      "loss": 2.2137,
+      "step": 4830
+    },
+    {
+      "epoch": 0.11398426828693892,
+      "grad_norm": 1.975380539894104,
+      "learning_rate": 0.00017721727662380483,
+      "loss": 2.2869,
+      "step": 4840
+    },
+    {
+      "epoch": 0.11421977297348217,
+      "grad_norm": 2.778428554534912,
+      "learning_rate": 0.00017717017568649615,
+      "loss": 2.151,
+      "step": 4850
+    },
+    {
+      "epoch": 0.11445527766002543,
+      "grad_norm": 2.3732733726501465,
+      "learning_rate": 0.00017712307474918753,
+      "loss": 2.1637,
+      "step": 4860
+    },
+    {
+      "epoch": 0.1146907823465687,
+      "grad_norm": 3.130542516708374,
+      "learning_rate": 0.00017707597381187885,
+      "loss": 2.2621,
+      "step": 4870
+    },
+    {
+      "epoch": 0.11492628703311196,
+      "grad_norm": 2.3386194705963135,
+      "learning_rate": 0.00017702887287457023,
+      "loss": 2.2108,
+      "step": 4880
+    },
+    {
+      "epoch": 0.11516179171965522,
+      "grad_norm": 1.9925601482391357,
+      "learning_rate": 0.00017698177193726155,
+      "loss": 2.2729,
+      "step": 4890
+    },
+    {
+      "epoch": 0.11539729640619849,
+      "grad_norm": 2.1639785766601562,
+      "learning_rate": 0.0001769346709999529,
+      "loss": 2.2079,
+      "step": 4900
+    },
+    {
+      "epoch": 0.11563280109274175,
+      "grad_norm": 2.6207916736602783,
+      "learning_rate": 0.00017688757006264425,
+      "loss": 2.1154,
+      "step": 4910
+    },
+    {
+      "epoch": 0.115868305779285,
+      "grad_norm": 2.0274581909179688,
+      "learning_rate": 0.0001768404691253356,
+      "loss": 2.2918,
+      "step": 4920
+    },
+    {
+      "epoch": 0.11610381046582827,
+      "grad_norm": 2.088710069656372,
+      "learning_rate": 0.00017679336818802695,
+      "loss": 2.2056,
+      "step": 4930
+    },
+    {
+      "epoch": 0.11633931515237153,
+      "grad_norm": 2.373544931411743,
+      "learning_rate": 0.0001767462672507183,
+      "loss": 2.1674,
+      "step": 4940
+    },
+    {
+      "epoch": 0.1165748198389148,
+      "grad_norm": 2.527215003967285,
+      "learning_rate": 0.00017669916631340963,
+      "loss": 1.9731,
+      "step": 4950
+    },
+    {
+      "epoch": 0.11681032452545806,
+      "grad_norm": 2.6474502086639404,
+      "learning_rate": 0.000176652065376101,
+      "loss": 2.0882,
+      "step": 4960
+    },
+    {
+      "epoch": 0.11704582921200132,
+      "grad_norm": 2.4764654636383057,
+      "learning_rate": 0.00017660496443879233,
+      "loss": 2.3867,
+      "step": 4970
+    },
+    {
+      "epoch": 0.11728133389854459,
+      "grad_norm": 1.8438410758972168,
+      "learning_rate": 0.00017655786350148368,
+      "loss": 2.3935,
+      "step": 4980
+    },
+    {
+      "epoch": 0.11751683858508784,
+      "grad_norm": 2.0502359867095947,
+      "learning_rate": 0.00017651076256417503,
+      "loss": 2.1566,
+      "step": 4990
+    },
+    {
+      "epoch": 0.1177523432716311,
+      "grad_norm": 2.705681562423706,
+      "learning_rate": 0.00017646366162686638,
+      "loss": 2.2442,
+      "step": 5000
+    },
+    {
+      "epoch": 0.11798784795817437,
+      "grad_norm": 1.7821354866027832,
+      "learning_rate": 0.00017641656068955773,
+      "loss": 2.095,
+      "step": 5010
+    },
+    {
+      "epoch": 0.11822335264471763,
+      "grad_norm": 2.171154022216797,
+      "learning_rate": 0.00017636945975224908,
+      "loss": 2.1128,
+      "step": 5020
+    },
+    {
+      "epoch": 0.1184588573312609,
+      "grad_norm": 1.8381446599960327,
+      "learning_rate": 0.00017632235881494043,
+      "loss": 2.2702,
+      "step": 5030
+    },
+    {
+      "epoch": 0.11869436201780416,
+      "grad_norm": 2.394317865371704,
+      "learning_rate": 0.00017627525787763178,
+      "loss": 2.2467,
+      "step": 5040
+    },
+    {
+      "epoch": 0.11892986670434742,
+      "grad_norm": 2.100402593612671,
+      "learning_rate": 0.00017622815694032313,
+      "loss": 2.1857,
+      "step": 5050
+    },
+    {
+      "epoch": 0.11916537139089067,
+      "grad_norm": 1.8596925735473633,
+      "learning_rate": 0.00017618105600301445,
+      "loss": 2.0607,
+      "step": 5060
+    },
+    {
+      "epoch": 0.11940087607743394,
+      "grad_norm": 2.523073196411133,
+      "learning_rate": 0.00017613395506570583,
+      "loss": 2.2643,
+      "step": 5070
+    },
+    {
+      "epoch": 0.1196363807639772,
+      "grad_norm": 1.6431152820587158,
+      "learning_rate": 0.00017608685412839715,
+      "loss": 2.3096,
+      "step": 5080
+    },
+    {
+      "epoch": 0.11987188545052047,
+      "grad_norm": 1.8515095710754395,
+      "learning_rate": 0.00017603975319108853,
+      "loss": 2.2366,
+      "step": 5090
+    },
+    {
+      "epoch": 0.12010739013706373,
+      "grad_norm": 2.4143874645233154,
+      "learning_rate": 0.00017599265225377985,
+      "loss": 2.2266,
+      "step": 5100
+    },
+    {
+      "epoch": 0.120342894823607,
+      "grad_norm": 2.4558234214782715,
+      "learning_rate": 0.0001759455513164712,
+      "loss": 2.3298,
+      "step": 5110
+    },
+    {
+      "epoch": 0.12057839951015026,
+      "grad_norm": 1.5276613235473633,
+      "learning_rate": 0.00017589845037916256,
+      "loss": 2.1767,
+      "step": 5120
+    },
+    {
+      "epoch": 0.12081390419669351,
+      "grad_norm": 2.6570372581481934,
+      "learning_rate": 0.0001758513494418539,
+      "loss": 2.3225,
+      "step": 5130
+    },
+    {
+      "epoch": 0.12104940888323677,
+      "grad_norm": 2.54841685295105,
+      "learning_rate": 0.00017580424850454526,
+      "loss": 2.3588,
+      "step": 5140
+    },
+    {
+      "epoch": 0.12128491356978004,
+      "grad_norm": 1.772416114807129,
+      "learning_rate": 0.0001757571475672366,
+      "loss": 2.3906,
+      "step": 5150
+    },
+    {
+      "epoch": 0.1215204182563233,
+      "grad_norm": 1.865113615989685,
+      "learning_rate": 0.00017571004662992793,
+      "loss": 2.2392,
+      "step": 5160
+    },
+    {
+      "epoch": 0.12175592294286657,
+      "grad_norm": 2.2344613075256348,
+      "learning_rate": 0.0001756629456926193,
+      "loss": 2.127,
+      "step": 5170
+    },
+    {
+      "epoch": 0.12199142762940983,
+      "grad_norm": 2.441580057144165,
+      "learning_rate": 0.00017561584475531063,
+      "loss": 2.2126,
+      "step": 5180
+    },
+    {
+      "epoch": 0.1222269323159531,
+      "grad_norm": 1.9913771152496338,
+      "learning_rate": 0.00017556874381800198,
+      "loss": 2.22,
+      "step": 5190
+    },
+    {
+      "epoch": 0.12246243700249634,
+      "grad_norm": 2.1991429328918457,
+      "learning_rate": 0.00017552164288069333,
+      "loss": 2.2766,
+      "step": 5200
+    },
+    {
+      "epoch": 0.12269794168903961,
+      "grad_norm": 2.120680809020996,
+      "learning_rate": 0.00017547454194338468,
+      "loss": 2.2078,
+      "step": 5210
+    },
+    {
+      "epoch": 0.12293344637558287,
+      "grad_norm": 1.788309931755066,
+      "learning_rate": 0.00017542744100607603,
+      "loss": 2.2587,
+      "step": 5220
+    },
+    {
+      "epoch": 0.12316895106212614,
+      "grad_norm": 2.2423155307769775,
+      "learning_rate": 0.00017538034006876738,
+      "loss": 2.2241,
+      "step": 5230
+    },
+    {
+      "epoch": 0.1234044557486694,
+      "grad_norm": 1.9412622451782227,
+      "learning_rate": 0.0001753332391314587,
+      "loss": 2.1715,
+      "step": 5240
+    },
+    {
+      "epoch": 0.12363996043521266,
+      "grad_norm": 2.0444014072418213,
+      "learning_rate": 0.00017528613819415008,
+      "loss": 2.1813,
+      "step": 5250
+    },
+    {
+      "epoch": 0.12387546512175593,
+      "grad_norm": 1.9657073020935059,
+      "learning_rate": 0.0001752390372568414,
+      "loss": 2.3767,
+      "step": 5260
+    },
+    {
+      "epoch": 0.12411096980829918,
+      "grad_norm": 2.156477928161621,
+      "learning_rate": 0.00017519193631953275,
+      "loss": 1.8883,
+      "step": 5270
+    },
+    {
+      "epoch": 0.12434647449484244,
+      "grad_norm": 2.735069513320923,
+      "learning_rate": 0.0001751448353822241,
+      "loss": 2.2131,
+      "step": 5280
+    },
+    {
+      "epoch": 0.12458197918138571,
+      "grad_norm": 1.98786199092865,
+      "learning_rate": 0.00017509773444491546,
+      "loss": 2.0419,
+      "step": 5290
+    },
+    {
+      "epoch": 0.12481748386792897,
+      "grad_norm": 2.619561195373535,
+      "learning_rate": 0.00017505063350760683,
+      "loss": 2.0047,
+      "step": 5300
+    },
+    {
+      "epoch": 0.12505298855447222,
+      "grad_norm": 2.2264404296875,
+      "learning_rate": 0.00017500353257029816,
+      "loss": 2.1834,
+      "step": 5310
+    },
+    {
+      "epoch": 0.1252884932410155,
+      "grad_norm": 2.1874210834503174,
+      "learning_rate": 0.0001749564316329895,
+      "loss": 2.1197,
+      "step": 5320
+    },
+    {
+      "epoch": 0.12552399792755875,
+      "grad_norm": 2.7099130153656006,
+      "learning_rate": 0.00017490933069568086,
+      "loss": 1.9282,
+      "step": 5330
+    },
+    {
+      "epoch": 0.12575950261410201,
+      "grad_norm": 2.1280713081359863,
+      "learning_rate": 0.0001748622297583722,
+      "loss": 2.0907,
+      "step": 5340
+    },
+    {
+      "epoch": 0.12599500730064528,
+      "grad_norm": 2.1866750717163086,
+      "learning_rate": 0.00017481512882106356,
+      "loss": 1.8259,
+      "step": 5350
+    },
+    {
+      "epoch": 0.12623051198718854,
+      "grad_norm": 1.9257513284683228,
+      "learning_rate": 0.0001747680278837549,
+      "loss": 2.1086,
+      "step": 5360
+    },
+    {
+      "epoch": 0.1264660166737318,
+      "grad_norm": 1.8964312076568604,
+      "learning_rate": 0.00017472092694644623,
+      "loss": 2.0995,
+      "step": 5370
+    },
+    {
+      "epoch": 0.12670152136027507,
+      "grad_norm": 2.228306770324707,
+      "learning_rate": 0.0001746738260091376,
+      "loss": 2.0373,
+      "step": 5380
+    },
+    {
+      "epoch": 0.12693702604681834,
+      "grad_norm": 2.723195791244507,
+      "learning_rate": 0.00017462672507182893,
+      "loss": 2.0848,
+      "step": 5390
+    },
+    {
+      "epoch": 0.1271725307333616,
+      "grad_norm": 2.1091506481170654,
+      "learning_rate": 0.00017457962413452028,
+      "loss": 2.3576,
+      "step": 5400
+    },
+    {
+      "epoch": 0.12740803541990486,
+      "grad_norm": 2.01425838470459,
+      "learning_rate": 0.00017453252319721163,
+      "loss": 2.1692,
+      "step": 5410
+    },
+    {
+      "epoch": 0.12764354010644813,
+      "grad_norm": 2.0158488750457764,
+      "learning_rate": 0.00017448542225990298,
+      "loss": 2.1845,
+      "step": 5420
+    },
+    {
+      "epoch": 0.1278790447929914,
+      "grad_norm": 1.796004056930542,
+      "learning_rate": 0.00017443832132259433,
+      "loss": 2.1808,
+      "step": 5430
+    },
+    {
+      "epoch": 0.12811454947953466,
+      "grad_norm": 2.268319845199585,
+      "learning_rate": 0.00017439122038528568,
+      "loss": 2.1978,
+      "step": 5440
+    },
+    {
+      "epoch": 0.1283500541660779,
+      "grad_norm": 2.573340654373169,
+      "learning_rate": 0.000174344119447977,
+      "loss": 2.2889,
+      "step": 5450
+    },
+    {
+      "epoch": 0.12858555885262116,
+      "grad_norm": 2.2684388160705566,
+      "learning_rate": 0.00017429701851066838,
+      "loss": 2.0348,
+      "step": 5460
+    },
+    {
+      "epoch": 0.12882106353916442,
+      "grad_norm": 2.271415948867798,
+      "learning_rate": 0.0001742499175733597,
+      "loss": 2.3534,
+      "step": 5470
+    },
+    {
+      "epoch": 0.12905656822570769,
+      "grad_norm": 1.900991439819336,
+      "learning_rate": 0.00017420281663605108,
+      "loss": 2.1018,
+      "step": 5480
+    },
+    {
+      "epoch": 0.12929207291225095,
+      "grad_norm": 2.262239456176758,
+      "learning_rate": 0.0001741557156987424,
+      "loss": 2.2768,
+      "step": 5490
+    },
+    {
+      "epoch": 0.1295275775987942,
+      "grad_norm": 2.041602849960327,
+      "learning_rate": 0.00017410861476143376,
+      "loss": 2.3032,
+      "step": 5500
+    },
+    {
+      "epoch": 0.12976308228533748,
+      "grad_norm": 1.4872256517410278,
+      "learning_rate": 0.0001740615138241251,
+      "loss": 2.3222,
+      "step": 5510
+    },
+    {
+      "epoch": 0.12999858697188074,
+      "grad_norm": 1.8539739847183228,
+      "learning_rate": 0.00017401441288681646,
+      "loss": 2.132,
+      "step": 5520
+    },
+    {
+      "epoch": 0.130234091658424,
+      "grad_norm": 2.243079900741577,
+      "learning_rate": 0.0001739673119495078,
+      "loss": 2.1824,
+      "step": 5530
+    },
+    {
+      "epoch": 0.13046959634496727,
+      "grad_norm": 2.0111191272735596,
+      "learning_rate": 0.00017392021101219916,
+      "loss": 2.1091,
+      "step": 5540
+    },
+    {
+      "epoch": 0.13070510103151053,
+      "grad_norm": 2.0866692066192627,
+      "learning_rate": 0.00017387311007489048,
+      "loss": 2.2156,
+      "step": 5550
+    },
+    {
+      "epoch": 0.1309406057180538,
+      "grad_norm": 1.6180267333984375,
+      "learning_rate": 0.00017382600913758186,
+      "loss": 2.2403,
+      "step": 5560
+    },
+    {
+      "epoch": 0.13117611040459706,
+      "grad_norm": 1.926936388015747,
+      "learning_rate": 0.00017378361829400408,
+      "loss": 2.268,
+      "step": 5570
+    },
+    {
+      "epoch": 0.13141161509114033,
+      "grad_norm": 1.732120156288147,
+      "learning_rate": 0.0001737365173566954,
+      "loss": 1.943,
+      "step": 5580
+    },
+    {
+      "epoch": 0.13164711977768356,
+      "grad_norm": 2.2308902740478516,
+      "learning_rate": 0.00017368941641938678,
+      "loss": 2.3538,
+      "step": 5590
+    },
+    {
+      "epoch": 0.13188262446422683,
+      "grad_norm": 2.0219004154205322,
+      "learning_rate": 0.0001736423154820781,
+      "loss": 2.1096,
+      "step": 5600
+    },
+    {
+      "epoch": 0.1321181291507701,
+      "grad_norm": 2.223156213760376,
+      "learning_rate": 0.00017359521454476945,
+      "loss": 2.138,
+      "step": 5610
+    },
+    {
+      "epoch": 0.13235363383731336,
+      "grad_norm": 2.6512413024902344,
+      "learning_rate": 0.0001735481136074608,
+      "loss": 1.9548,
+      "step": 5620
+    },
+    {
+      "epoch": 0.13258913852385662,
+      "grad_norm": 1.974619746208191,
+      "learning_rate": 0.00017350101267015215,
+      "loss": 2.1642,
+      "step": 5630
+    },
+    {
+      "epoch": 0.13282464321039988,
+      "grad_norm": 2.279487133026123,
+      "learning_rate": 0.0001734539117328435,
+      "loss": 2.1445,
+      "step": 5640
+    },
+    {
+      "epoch": 0.13306014789694315,
+      "grad_norm": 1.9778087139129639,
+      "learning_rate": 0.00017340681079553485,
+      "loss": 2.2169,
+      "step": 5650
+    },
+    {
+      "epoch": 0.1332956525834864,
+      "grad_norm": 1.9893101453781128,
+      "learning_rate": 0.00017335970985822617,
+      "loss": 2.2353,
+      "step": 5660
+    },
+    {
+      "epoch": 0.13353115727002968,
+      "grad_norm": 2.7806546688079834,
+      "learning_rate": 0.00017331260892091755,
+      "loss": 2.1214,
+      "step": 5670
+    },
+    {
+      "epoch": 0.13376666195657294,
+      "grad_norm": 2.2366230487823486,
+      "learning_rate": 0.00017326550798360887,
+      "loss": 2.1004,
+      "step": 5680
+    },
+    {
+      "epoch": 0.1340021666431162,
+      "grad_norm": 3.1434006690979004,
+      "learning_rate": 0.00017321840704630022,
+      "loss": 2.3682,
+      "step": 5690
+    },
+    {
+      "epoch": 0.13423767132965947,
+      "grad_norm": 1.862724781036377,
+      "learning_rate": 0.00017317130610899157,
+      "loss": 2.1701,
+      "step": 5700
+    },
+    {
+      "epoch": 0.13447317601620273,
+      "grad_norm": 2.2287230491638184,
+      "learning_rate": 0.00017312420517168292,
+      "loss": 2.2376,
+      "step": 5710
+    },
+    {
+      "epoch": 0.134708680702746,
+      "grad_norm": 2.0166916847229004,
+      "learning_rate": 0.00017307710423437428,
+      "loss": 2.2481,
+      "step": 5720
+    },
+    {
+      "epoch": 0.13494418538928923,
+      "grad_norm": 2.4379286766052246,
+      "learning_rate": 0.00017303000329706563,
+      "loss": 2.2884,
+      "step": 5730
+    },
+    {
+      "epoch": 0.1351796900758325,
+      "grad_norm": 1.923784613609314,
+      "learning_rate": 0.00017298290235975695,
+      "loss": 2.13,
+      "step": 5740
+    },
+    {
+      "epoch": 0.13541519476237576,
+      "grad_norm": 2.3278648853302,
+      "learning_rate": 0.00017293580142244833,
+      "loss": 2.4975,
+      "step": 5750
+    },
+    {
+      "epoch": 0.13565069944891903,
+      "grad_norm": 2.186098098754883,
+      "learning_rate": 0.00017288870048513965,
+      "loss": 2.4386,
+      "step": 5760
+    },
+    {
+      "epoch": 0.1358862041354623,
+      "grad_norm": 1.8894046545028687,
+      "learning_rate": 0.000172841599547831,
+      "loss": 2.1392,
+      "step": 5770
+    },
+    {
+      "epoch": 0.13612170882200555,
+      "grad_norm": 2.2432827949523926,
+      "learning_rate": 0.00017279449861052235,
+      "loss": 2.2693,
+      "step": 5780
+    },
+    {
+      "epoch": 0.13635721350854882,
+      "grad_norm": 1.860554575920105,
+      "learning_rate": 0.0001727473976732137,
+      "loss": 2.0159,
+      "step": 5790
+    },
+    {
+      "epoch": 0.13659271819509208,
+      "grad_norm": 1.6823490858078003,
+      "learning_rate": 0.00017270029673590505,
+      "loss": 2.2196,
+      "step": 5800
+    },
+    {
+      "epoch": 0.13682822288163535,
+      "grad_norm": 1.9365490674972534,
+      "learning_rate": 0.0001726531957985964,
+      "loss": 2.2016,
+      "step": 5810
+    },
+    {
+      "epoch": 0.1370637275681786,
+      "grad_norm": 2.2856075763702393,
+      "learning_rate": 0.00017260609486128775,
+      "loss": 2.2524,
+      "step": 5820
+    },
+    {
+      "epoch": 0.13729923225472188,
+      "grad_norm": 2.667853832244873,
+      "learning_rate": 0.0001725589939239791,
+      "loss": 1.9745,
+      "step": 5830
+    },
+    {
+      "epoch": 0.13753473694126514,
+      "grad_norm": 1.9969861507415771,
+      "learning_rate": 0.00017251189298667045,
+      "loss": 2.1179,
+      "step": 5840
+    },
+    {
+      "epoch": 0.1377702416278084,
+      "grad_norm": 1.866898536682129,
+      "learning_rate": 0.0001724647920493618,
+      "loss": 2.3985,
+      "step": 5850
+    },
+    {
+      "epoch": 0.13800574631435167,
+      "grad_norm": 2.4066002368927,
+      "learning_rate": 0.00017241769111205315,
+      "loss": 2.2191,
+      "step": 5860
+    },
+    {
+      "epoch": 0.1382412510008949,
+      "grad_norm": 2.1669511795043945,
+      "learning_rate": 0.00017237059017474447,
+      "loss": 2.07,
+      "step": 5870
+    },
+    {
+      "epoch": 0.13847675568743817,
+      "grad_norm": 1.9920525550842285,
+      "learning_rate": 0.00017232348923743585,
+      "loss": 2.1086,
+      "step": 5880
+    },
+    {
+      "epoch": 0.13871226037398143,
+      "grad_norm": 2.622621536254883,
+      "learning_rate": 0.00017227638830012717,
+      "loss": 2.1982,
+      "step": 5890
+    },
+    {
+      "epoch": 0.1389477650605247,
+      "grad_norm": 2.285757064819336,
+      "learning_rate": 0.00017222928736281853,
+      "loss": 2.2294,
+      "step": 5900
+    },
+    {
+      "epoch": 0.13918326974706796,
+      "grad_norm": 2.428218126296997,
+      "learning_rate": 0.00017218218642550988,
+      "loss": 2.3078,
+      "step": 5910
+    },
+    {
+      "epoch": 0.13941877443361123,
+      "grad_norm": 1.8226295709609985,
+      "learning_rate": 0.00017213508548820123,
+      "loss": 2.1509,
+      "step": 5920
+    },
+    {
+      "epoch": 0.1396542791201545,
+      "grad_norm": 2.660916328430176,
+      "learning_rate": 0.00017208798455089258,
+      "loss": 2.0719,
+      "step": 5930
+    },
+    {
+      "epoch": 0.13988978380669775,
+      "grad_norm": 1.9851837158203125,
+      "learning_rate": 0.00017204088361358393,
+      "loss": 2.0906,
+      "step": 5940
+    },
+    {
+      "epoch": 0.14012528849324102,
+      "grad_norm": 2.3256642818450928,
+      "learning_rate": 0.00017199378267627525,
+      "loss": 2.0444,
+      "step": 5950
+    },
+    {
+      "epoch": 0.14036079317978428,
+      "grad_norm": 1.7920613288879395,
+      "learning_rate": 0.00017194668173896663,
+      "loss": 2.1493,
+      "step": 5960
+    },
+    {
+      "epoch": 0.14059629786632755,
+      "grad_norm": 1.760733723640442,
+      "learning_rate": 0.00017189958080165795,
+      "loss": 2.3474,
+      "step": 5970
+    },
+    {
+      "epoch": 0.1408318025528708,
+      "grad_norm": 2.066558599472046,
+      "learning_rate": 0.00017185247986434933,
+      "loss": 2.1801,
+      "step": 5980
+    },
+    {
+      "epoch": 0.14106730723941407,
+      "grad_norm": 1.9565317630767822,
+      "learning_rate": 0.00017180537892704065,
+      "loss": 2.3232,
+      "step": 5990
+    },
+    {
+      "epoch": 0.14130281192595734,
+      "grad_norm": 2.2794132232666016,
+      "learning_rate": 0.000171758277989732,
+      "loss": 2.178,
+      "step": 6000
+    },
+    {
+      "epoch": 0.14153831661250058,
+      "grad_norm": 1.8281110525131226,
+      "learning_rate": 0.00017171117705242335,
+      "loss": 2.1067,
+      "step": 6010
+    },
+    {
+      "epoch": 0.14177382129904384,
+      "grad_norm": 1.9072505235671997,
+      "learning_rate": 0.0001716640761151147,
+      "loss": 1.9927,
+      "step": 6020
+    },
+    {
+      "epoch": 0.1420093259855871,
+      "grad_norm": 2.288691997528076,
+      "learning_rate": 0.00017161697517780605,
+      "loss": 2.0077,
+      "step": 6030
+    },
+    {
+      "epoch": 0.14224483067213037,
+      "grad_norm": 2.0994765758514404,
+      "learning_rate": 0.0001715698742404974,
+      "loss": 2.2116,
+      "step": 6040
+    },
+    {
+      "epoch": 0.14248033535867363,
+      "grad_norm": 1.807565689086914,
+      "learning_rate": 0.00017152277330318872,
+      "loss": 2.3081,
+      "step": 6050
+    },
+    {
+      "epoch": 0.1427158400452169,
+      "grad_norm": 1.8691102266311646,
+      "learning_rate": 0.0001714756723658801,
+      "loss": 2.1234,
+      "step": 6060
+    },
+    {
+      "epoch": 0.14295134473176016,
+      "grad_norm": 2.168907403945923,
+      "learning_rate": 0.00017142857142857143,
+      "loss": 2.1899,
+      "step": 6070
+    },
+    {
+      "epoch": 0.14318684941830342,
+      "grad_norm": 2.2471699714660645,
+      "learning_rate": 0.00017138147049126278,
+      "loss": 2.3476,
+      "step": 6080
+    },
+    {
+      "epoch": 0.1434223541048467,
+      "grad_norm": 1.8426363468170166,
+      "learning_rate": 0.00017133436955395413,
+      "loss": 2.0467,
+      "step": 6090
+    },
+    {
+      "epoch": 0.14365785879138995,
+      "grad_norm": 1.6236611604690552,
+      "learning_rate": 0.00017128726861664548,
+      "loss": 2.0523,
+      "step": 6100
+    },
+    {
+      "epoch": 0.14389336347793322,
+      "grad_norm": 1.660072922706604,
+      "learning_rate": 0.00017124016767933683,
+      "loss": 2.2243,
+      "step": 6110
+    },
+    {
+      "epoch": 0.14412886816447648,
+      "grad_norm": 2.4543228149414062,
+      "learning_rate": 0.00017119306674202818,
+      "loss": 2.1948,
+      "step": 6120
+    },
+    {
+      "epoch": 0.14436437285101975,
+      "grad_norm": 1.9436759948730469,
+      "learning_rate": 0.00017114596580471953,
+      "loss": 2.0913,
+      "step": 6130
+    },
+    {
+      "epoch": 0.144599877537563,
+      "grad_norm": 2.1361687183380127,
+      "learning_rate": 0.00017109886486741088,
+      "loss": 2.2386,
+      "step": 6140
+    },
+    {
+      "epoch": 0.14483538222410625,
+      "grad_norm": 3.258145332336426,
+      "learning_rate": 0.00017105176393010223,
+      "loss": 2.1049,
+      "step": 6150
+    },
+    {
+      "epoch": 0.1450708869106495,
+      "grad_norm": 2.3219218254089355,
+      "learning_rate": 0.00017100466299279355,
+      "loss": 2.2702,
+      "step": 6160
+    },
+    {
+      "epoch": 0.14530639159719277,
+      "grad_norm": 2.53751277923584,
+      "learning_rate": 0.00017095756205548493,
+      "loss": 2.1055,
+      "step": 6170
+    },
+    {
+      "epoch": 0.14554189628373604,
+      "grad_norm": 1.9522231817245483,
+      "learning_rate": 0.00017091046111817625,
+      "loss": 2.1196,
+      "step": 6180
+    },
+    {
+      "epoch": 0.1457774009702793,
+      "grad_norm": 1.9058953523635864,
+      "learning_rate": 0.00017086336018086763,
+      "loss": 2.1288,
+      "step": 6190
+    },
+    {
+      "epoch": 0.14601290565682257,
+      "grad_norm": 2.3197526931762695,
+      "learning_rate": 0.00017081625924355895,
+      "loss": 2.0571,
+      "step": 6200
+    },
+    {
+      "epoch": 0.14624841034336583,
+      "grad_norm": 2.9913275241851807,
+      "learning_rate": 0.0001707691583062503,
+      "loss": 2.042,
+      "step": 6210
+    },
+    {
+      "epoch": 0.1464839150299091,
+      "grad_norm": 3.6737120151519775,
+      "learning_rate": 0.00017072205736894165,
+      "loss": 2.09,
+      "step": 6220
+    },
+    {
+      "epoch": 0.14671941971645236,
+      "grad_norm": 2.2069551944732666,
+      "learning_rate": 0.000170674956431633,
+      "loss": 2.1867,
+      "step": 6230
+    },
+    {
+      "epoch": 0.14695492440299562,
+      "grad_norm": 2.2664968967437744,
+      "learning_rate": 0.00017062785549432435,
+      "loss": 2.2181,
+      "step": 6240
+    },
+    {
+      "epoch": 0.1471904290895389,
+      "grad_norm": 1.966537356376648,
+      "learning_rate": 0.0001705807545570157,
+      "loss": 2.0683,
+      "step": 6250
+    },
+    {
+      "epoch": 0.14742593377608215,
+      "grad_norm": 1.8831692934036255,
+      "learning_rate": 0.00017053365361970703,
+      "loss": 2.0964,
+      "step": 6260
+    },
+    {
+      "epoch": 0.14766143846262542,
+      "grad_norm": 2.0618419647216797,
+      "learning_rate": 0.0001704865526823984,
+      "loss": 2.3007,
+      "step": 6270
+    },
+    {
+      "epoch": 0.14789694314916868,
+      "grad_norm": 2.163808584213257,
+      "learning_rate": 0.00017043945174508973,
+      "loss": 2.1509,
+      "step": 6280
+    },
+    {
+      "epoch": 0.14813244783571194,
+      "grad_norm": 2.259477138519287,
+      "learning_rate": 0.00017039235080778108,
+      "loss": 2.2863,
+      "step": 6290
+    },
+    {
+      "epoch": 0.14836795252225518,
+      "grad_norm": 1.4202922582626343,
+      "learning_rate": 0.00017034524987047243,
+      "loss": 1.9795,
+      "step": 6300
+    },
+    {
+      "epoch": 0.14860345720879845,
+      "grad_norm": 2.2261738777160645,
+      "learning_rate": 0.00017029814893316378,
+      "loss": 2.0105,
+      "step": 6310
+    },
+    {
+      "epoch": 0.1488389618953417,
+      "grad_norm": 2.284642457962036,
+      "learning_rate": 0.00017025104799585513,
+      "loss": 2.1962,
+      "step": 6320
+    },
+    {
+      "epoch": 0.14907446658188497,
+      "grad_norm": 1.6194367408752441,
+      "learning_rate": 0.00017020394705854648,
+      "loss": 2.0346,
+      "step": 6330
+    },
+    {
+      "epoch": 0.14930997126842824,
+      "grad_norm": 2.0232205390930176,
+      "learning_rate": 0.0001701568461212378,
+      "loss": 2.2708,
+      "step": 6340
+    },
+    {
+      "epoch": 0.1495454759549715,
+      "grad_norm": 2.325477361679077,
+      "learning_rate": 0.00017010974518392918,
+      "loss": 2.1813,
+      "step": 6350
+    },
+    {
+      "epoch": 0.14978098064151477,
+      "grad_norm": 2.309537172317505,
+      "learning_rate": 0.0001700626442466205,
+      "loss": 2.1677,
+      "step": 6360
+    },
+    {
+      "epoch": 0.15001648532805803,
+      "grad_norm": 2.201800584793091,
+      "learning_rate": 0.00017001554330931185,
+      "loss": 2.1772,
+      "step": 6370
+    },
+    {
+      "epoch": 0.1502519900146013,
+      "grad_norm": 1.840621829032898,
+      "learning_rate": 0.0001699684423720032,
+      "loss": 2.1918,
+      "step": 6380
+    },
+    {
+      "epoch": 0.15048749470114456,
+      "grad_norm": 2.601182222366333,
+      "learning_rate": 0.00016992134143469455,
+      "loss": 2.1371,
+      "step": 6390
+    },
+    {
+      "epoch": 0.15072299938768782,
+      "grad_norm": 1.4714235067367554,
+      "learning_rate": 0.00016987424049738593,
+      "loss": 2.0602,
+      "step": 6400
+    },
+    {
+      "epoch": 0.1509585040742311,
+      "grad_norm": 2.3133628368377686,
+      "learning_rate": 0.00016982713956007725,
+      "loss": 2.3063,
+      "step": 6410
+    },
+    {
+      "epoch": 0.15119400876077435,
+      "grad_norm": 1.9627376794815063,
+      "learning_rate": 0.0001697800386227686,
+      "loss": 1.9214,
+      "step": 6420
+    },
+    {
+      "epoch": 0.15142951344731762,
+      "grad_norm": 2.3527040481567383,
+      "learning_rate": 0.00016973293768545995,
+      "loss": 2.1646,
+      "step": 6430
+    },
+    {
+      "epoch": 0.15166501813386085,
+      "grad_norm": 2.2275426387786865,
+      "learning_rate": 0.0001696858367481513,
+      "loss": 2.1793,
+      "step": 6440
+    },
+    {
+      "epoch": 0.15190052282040412,
+      "grad_norm": 2.1824982166290283,
+      "learning_rate": 0.00016963873581084265,
+      "loss": 2.2669,
+      "step": 6450
+    },
+    {
+      "epoch": 0.15213602750694738,
+      "grad_norm": 2.512240171432495,
+      "learning_rate": 0.000169591634873534,
+      "loss": 2.1601,
+      "step": 6460
+    },
+    {
+      "epoch": 0.15237153219349064,
+      "grad_norm": 2.1347885131835938,
+      "learning_rate": 0.00016954453393622533,
+      "loss": 2.1572,
+      "step": 6470
+    },
+    {
+      "epoch": 0.1526070368800339,
+      "grad_norm": 1.7296582460403442,
+      "learning_rate": 0.0001694974329989167,
+      "loss": 2.2337,
+      "step": 6480
+    },
+    {
+      "epoch": 0.15284254156657717,
+      "grad_norm": 2.2184410095214844,
+      "learning_rate": 0.00016945033206160803,
+      "loss": 2.1873,
+      "step": 6490
+    },
+    {
+      "epoch": 0.15307804625312044,
+      "grad_norm": 1.787880778312683,
+      "learning_rate": 0.00016940323112429938,
+      "loss": 2.0755,
+      "step": 6500
+    },
+    {
+      "epoch": 0.1533135509396637,
+      "grad_norm": 1.9352377653121948,
+      "learning_rate": 0.00016935613018699073,
+      "loss": 2.2166,
+      "step": 6510
+    },
+    {
+      "epoch": 0.15354905562620697,
+      "grad_norm": 2.27699875831604,
+      "learning_rate": 0.00016930902924968208,
+      "loss": 2.2189,
+      "step": 6520
+    },
+    {
+      "epoch": 0.15378456031275023,
+      "grad_norm": 1.4749678373336792,
+      "learning_rate": 0.00016926192831237343,
+      "loss": 1.9943,
+      "step": 6530
+    },
+    {
+      "epoch": 0.1540200649992935,
+      "grad_norm": 2.7707648277282715,
+      "learning_rate": 0.00016921482737506478,
+      "loss": 2.3423,
+      "step": 6540
+    },
+    {
+      "epoch": 0.15425556968583676,
+      "grad_norm": 2.5478742122650146,
+      "learning_rate": 0.0001691677264377561,
+      "loss": 2.1751,
+      "step": 6550
+    },
+    {
+      "epoch": 0.15449107437238002,
+      "grad_norm": 2.3068339824676514,
+      "learning_rate": 0.00016912062550044748,
+      "loss": 2.2444,
+      "step": 6560
+    },
+    {
+      "epoch": 0.15472657905892329,
+      "grad_norm": 1.8547791242599487,
+      "learning_rate": 0.0001690735245631388,
+      "loss": 2.436,
+      "step": 6570
+    },
+    {
+      "epoch": 0.15496208374546652,
+      "grad_norm": 2.1598117351531982,
+      "learning_rate": 0.00016902642362583018,
+      "loss": 2.0398,
+      "step": 6580
+    },
+    {
+      "epoch": 0.1551975884320098,
+      "grad_norm": 2.000115156173706,
+      "learning_rate": 0.0001689793226885215,
+      "loss": 1.9789,
+      "step": 6590
+    },
+    {
+      "epoch": 0.15543309311855305,
+      "grad_norm": 2.3992674350738525,
+      "learning_rate": 0.00016893222175121285,
+      "loss": 2.2092,
+      "step": 6600
+    },
+    {
+      "epoch": 0.15566859780509631,
+      "grad_norm": 2.2532436847686768,
+      "learning_rate": 0.0001688851208139042,
+      "loss": 2.0844,
+      "step": 6610
+    },
+    {
+      "epoch": 0.15590410249163958,
+      "grad_norm": 1.8847017288208008,
+      "learning_rate": 0.00016883801987659555,
+      "loss": 2.0509,
+      "step": 6620
+    },
+    {
+      "epoch": 0.15613960717818284,
+      "grad_norm": 2.4338490962982178,
+      "learning_rate": 0.0001687909189392869,
+      "loss": 2.229,
+      "step": 6630
+    },
+    {
+      "epoch": 0.1563751118647261,
+      "grad_norm": 1.6963534355163574,
+      "learning_rate": 0.00016874381800197825,
+      "loss": 2.3512,
+      "step": 6640
+    },
+    {
+      "epoch": 0.15661061655126937,
+      "grad_norm": 2.159989356994629,
+      "learning_rate": 0.00016869671706466958,
+      "loss": 2.1962,
+      "step": 6650
+    },
+    {
+      "epoch": 0.15684612123781264,
+      "grad_norm": 2.163346290588379,
+      "learning_rate": 0.00016864961612736095,
+      "loss": 2.2327,
+      "step": 6660
+    },
+    {
+      "epoch": 0.1570816259243559,
+      "grad_norm": 3.266700029373169,
+      "learning_rate": 0.00016860251519005228,
+      "loss": 2.0636,
+      "step": 6670
+    },
+    {
+      "epoch": 0.15731713061089916,
+      "grad_norm": 2.0337531566619873,
+      "learning_rate": 0.00016855541425274363,
+      "loss": 2.2385,
+      "step": 6680
+    },
+    {
+      "epoch": 0.15755263529744243,
+      "grad_norm": 2.7667040824890137,
+      "learning_rate": 0.000168508313315435,
+      "loss": 2.1956,
+      "step": 6690
+    },
+    {
+      "epoch": 0.1577881399839857,
+      "grad_norm": 2.1070220470428467,
+      "learning_rate": 0.00016846121237812633,
+      "loss": 2.1798,
+      "step": 6700
+    },
+    {
+      "epoch": 0.15802364467052896,
+      "grad_norm": 1.6910346746444702,
+      "learning_rate": 0.00016841411144081768,
+      "loss": 2.1594,
+      "step": 6710
+    },
+    {
+      "epoch": 0.1582591493570722,
+      "grad_norm": 2.147885799407959,
+      "learning_rate": 0.00016836701050350903,
+      "loss": 2.2545,
+      "step": 6720
+    },
+    {
+      "epoch": 0.15849465404361546,
+      "grad_norm": 2.54560923576355,
+      "learning_rate": 0.00016831990956620038,
+      "loss": 2.0461,
+      "step": 6730
+    },
+    {
+      "epoch": 0.15873015873015872,
+      "grad_norm": 2.0715444087982178,
+      "learning_rate": 0.00016827280862889173,
+      "loss": 2.3273,
+      "step": 6740
+    },
+    {
+      "epoch": 0.15896566341670199,
+      "grad_norm": 2.0344667434692383,
+      "learning_rate": 0.00016822570769158308,
+      "loss": 2.2495,
+      "step": 6750
+    },
+    {
+      "epoch": 0.15920116810324525,
+      "grad_norm": 2.010301113128662,
+      "learning_rate": 0.0001681786067542744,
+      "loss": 2.2644,
+      "step": 6760
+    },
+    {
+      "epoch": 0.1594366727897885,
+      "grad_norm": 2.423848867416382,
+      "learning_rate": 0.00016813150581696578,
+      "loss": 1.9635,
+      "step": 6770
+    },
+    {
+      "epoch": 0.15967217747633178,
+      "grad_norm": 1.817819595336914,
+      "learning_rate": 0.0001680844048796571,
+      "loss": 2.1507,
+      "step": 6780
+    },
+    {
+      "epoch": 0.15990768216287504,
+      "grad_norm": 2.4440290927886963,
+      "learning_rate": 0.00016803730394234848,
+      "loss": 2.641,
+      "step": 6790
+    },
+    {
+      "epoch": 0.1601431868494183,
+      "grad_norm": 1.7758880853652954,
+      "learning_rate": 0.0001679902030050398,
+      "loss": 2.0812,
+      "step": 6800
+    },
+    {
+      "epoch": 0.16037869153596157,
+      "grad_norm": 2.6482930183410645,
+      "learning_rate": 0.00016794310206773115,
+      "loss": 2.0167,
+      "step": 6810
+    },
+    {
+      "epoch": 0.16061419622250483,
+      "grad_norm": 2.0901808738708496,
+      "learning_rate": 0.0001678960011304225,
+      "loss": 1.9824,
+      "step": 6820
+    },
+    {
+      "epoch": 0.1608497009090481,
+      "grad_norm": 2.458341121673584,
+      "learning_rate": 0.00016784890019311385,
+      "loss": 2.0419,
+      "step": 6830
+    },
+    {
+      "epoch": 0.16108520559559136,
+      "grad_norm": 1.7757121324539185,
+      "learning_rate": 0.0001678017992558052,
+      "loss": 2.1314,
+      "step": 6840
+    },
+    {
+      "epoch": 0.16132071028213463,
+      "grad_norm": 2.2329952716827393,
+      "learning_rate": 0.00016775469831849655,
+      "loss": 2.0286,
+      "step": 6850
+    },
+    {
+      "epoch": 0.16155621496867786,
+      "grad_norm": 2.1423516273498535,
+      "learning_rate": 0.00016770759738118788,
+      "loss": 2.21,
+      "step": 6860
+    },
+    {
+      "epoch": 0.16179171965522113,
+      "grad_norm": 1.869727611541748,
+      "learning_rate": 0.00016766049644387925,
+      "loss": 2.0268,
+      "step": 6870
+    },
+    {
+      "epoch": 0.1620272243417644,
+      "grad_norm": 2.514774799346924,
+      "learning_rate": 0.00016761339550657058,
+      "loss": 2.164,
+      "step": 6880
+    },
+    {
+      "epoch": 0.16226272902830766,
+      "grad_norm": 1.9911197423934937,
+      "learning_rate": 0.00016756629456926193,
+      "loss": 2.0608,
+      "step": 6890
+    },
+    {
+      "epoch": 0.16249823371485092,
+      "grad_norm": 1.9480197429656982,
+      "learning_rate": 0.00016751919363195328,
+      "loss": 2.1584,
+      "step": 6900
+    },
+    {
+      "epoch": 0.16273373840139418,
+      "grad_norm": 3.6419789791107178,
+      "learning_rate": 0.00016747209269464463,
+      "loss": 2.0359,
+      "step": 6910
+    },
+    {
+      "epoch": 0.16296924308793745,
+      "grad_norm": 1.879197597503662,
+      "learning_rate": 0.00016742499175733598,
+      "loss": 1.9777,
+      "step": 6920
+    },
+    {
+      "epoch": 0.1632047477744807,
+      "grad_norm": 2.217308759689331,
+      "learning_rate": 0.00016737789082002733,
+      "loss": 2.3454,
+      "step": 6930
+    },
+    {
+      "epoch": 0.16344025246102398,
+      "grad_norm": 2.226469039916992,
+      "learning_rate": 0.00016733078988271865,
+      "loss": 2.1253,
+      "step": 6940
+    },
+    {
+      "epoch": 0.16367575714756724,
+      "grad_norm": 2.5227324962615967,
+      "learning_rate": 0.00016728368894541003,
+      "loss": 2.0713,
+      "step": 6950
+    },
+    {
+      "epoch": 0.1639112618341105,
+      "grad_norm": 1.8012442588806152,
+      "learning_rate": 0.00016723658800810138,
+      "loss": 1.9799,
+      "step": 6960
+    },
+    {
+      "epoch": 0.16414676652065377,
+      "grad_norm": 3.06243634223938,
+      "learning_rate": 0.0001671894870707927,
+      "loss": 2.3178,
+      "step": 6970
+    },
+    {
+      "epoch": 0.16438227120719703,
+      "grad_norm": 1.4060784578323364,
+      "learning_rate": 0.00016714238613348408,
+      "loss": 2.149,
+      "step": 6980
+    },
+    {
+      "epoch": 0.1646177758937403,
+      "grad_norm": 1.7966259717941284,
+      "learning_rate": 0.0001670952851961754,
+      "loss": 2.2267,
+      "step": 6990
+    },
+    {
+      "epoch": 0.16485328058028353,
+      "grad_norm": 1.7830440998077393,
+      "learning_rate": 0.00016704818425886678,
+      "loss": 2.1592,
+      "step": 7000
+    },
+    {
+      "epoch": 0.1650887852668268,
+      "grad_norm": 1.8834739923477173,
+      "learning_rate": 0.0001670010833215581,
+      "loss": 2.0868,
+      "step": 7010
+    },
+    {
+      "epoch": 0.16532428995337006,
+      "grad_norm": 2.2466180324554443,
+      "learning_rate": 0.00016695398238424945,
+      "loss": 2.1263,
+      "step": 7020
+    },
+    {
+      "epoch": 0.16555979463991333,
+      "grad_norm": 2.24786639213562,
+      "learning_rate": 0.0001669068814469408,
+      "loss": 2.0783,
+      "step": 7030
+    },
+    {
+      "epoch": 0.1657952993264566,
+      "grad_norm": 2.1359407901763916,
+      "learning_rate": 0.00016685978050963215,
+      "loss": 2.2205,
+      "step": 7040
+    },
+    {
+      "epoch": 0.16603080401299986,
+      "grad_norm": 1.937888741493225,
+      "learning_rate": 0.0001668126795723235,
+      "loss": 2.191,
+      "step": 7050
+    },
+    {
+      "epoch": 0.16626630869954312,
+      "grad_norm": 2.25390625,
+      "learning_rate": 0.00016676557863501486,
+      "loss": 2.0718,
+      "step": 7060
+    },
+    {
+      "epoch": 0.16650181338608638,
+      "grad_norm": 1.8093719482421875,
+      "learning_rate": 0.00016671847769770618,
+      "loss": 2.133,
+      "step": 7070
+    },
+    {
+      "epoch": 0.16673731807262965,
+      "grad_norm": 2.38726806640625,
+      "learning_rate": 0.00016667137676039756,
+      "loss": 2.2872,
+      "step": 7080
+    },
+    {
+      "epoch": 0.1669728227591729,
+      "grad_norm": 1.934637188911438,
+      "learning_rate": 0.00016662427582308888,
+      "loss": 2.3178,
+      "step": 7090
+    },
+    {
+      "epoch": 0.16720832744571618,
+      "grad_norm": 2.438570976257324,
+      "learning_rate": 0.00016657717488578023,
+      "loss": 1.9654,
+      "step": 7100
+    },
+    {
+      "epoch": 0.16744383213225944,
+      "grad_norm": 2.067674398422241,
+      "learning_rate": 0.00016653007394847158,
+      "loss": 2.2299,
+      "step": 7110
+    },
+    {
+      "epoch": 0.1676793368188027,
+      "grad_norm": 2.227522373199463,
+      "learning_rate": 0.00016648297301116293,
+      "loss": 2.2692,
+      "step": 7120
+    },
+    {
+      "epoch": 0.16791484150534597,
+      "grad_norm": 4.050379753112793,
+      "learning_rate": 0.00016643587207385428,
+      "loss": 2.2243,
+      "step": 7130
+    },
+    {
+      "epoch": 0.1681503461918892,
+      "grad_norm": 2.220202922821045,
+      "learning_rate": 0.00016638877113654563,
+      "loss": 2.2231,
+      "step": 7140
+    },
+    {
+      "epoch": 0.16838585087843247,
+      "grad_norm": 2.7050681114196777,
+      "learning_rate": 0.00016634167019923695,
+      "loss": 2.1957,
+      "step": 7150
+    },
+    {
+      "epoch": 0.16862135556497573,
+      "grad_norm": 2.3848469257354736,
+      "learning_rate": 0.00016629456926192833,
+      "loss": 2.0928,
+      "step": 7160
+    },
+    {
+      "epoch": 0.168856860251519,
+      "grad_norm": 2.007478713989258,
+      "learning_rate": 0.00016624746832461965,
+      "loss": 2.1006,
+      "step": 7170
+    },
+    {
+      "epoch": 0.16909236493806226,
+      "grad_norm": 2.2802834510803223,
+      "learning_rate": 0.00016620036738731103,
+      "loss": 2.2195,
+      "step": 7180
+    },
+    {
+      "epoch": 0.16932786962460553,
+      "grad_norm": 2.2520174980163574,
+      "learning_rate": 0.00016615326645000235,
+      "loss": 2.169,
+      "step": 7190
+    },
+    {
+      "epoch": 0.1695633743111488,
+      "grad_norm": 2.1334917545318604,
+      "learning_rate": 0.0001661061655126937,
+      "loss": 2.3443,
+      "step": 7200
+    },
+    {
+      "epoch": 0.16979887899769205,
+      "grad_norm": 2.5005075931549072,
+      "learning_rate": 0.00016605906457538505,
+      "loss": 2.1891,
+      "step": 7210
+    },
+    {
+      "epoch": 0.17003438368423532,
+      "grad_norm": 2.019739866256714,
+      "learning_rate": 0.0001660119636380764,
+      "loss": 2.1287,
+      "step": 7220
+    },
+    {
+      "epoch": 0.17026988837077858,
+      "grad_norm": 1.7916017770767212,
+      "learning_rate": 0.00016596486270076776,
+      "loss": 2.1305,
+      "step": 7230
+    },
+    {
+      "epoch": 0.17050539305732185,
+      "grad_norm": 1.9927102327346802,
+      "learning_rate": 0.0001659177617634591,
+      "loss": 2.0284,
+      "step": 7240
+    },
+    {
+      "epoch": 0.1707408977438651,
+      "grad_norm": 2.6178619861602783,
+      "learning_rate": 0.00016587066082615046,
+      "loss": 2.2635,
+      "step": 7250
+    },
+    {
+      "epoch": 0.17097640243040838,
+      "grad_norm": 2.3916609287261963,
+      "learning_rate": 0.0001658235598888418,
+      "loss": 2.243,
+      "step": 7260
+    },
+    {
+      "epoch": 0.17121190711695164,
+      "grad_norm": 2.027822494506836,
+      "learning_rate": 0.00016577645895153316,
+      "loss": 2.1836,
+      "step": 7270
+    },
+    {
+      "epoch": 0.17144741180349488,
+      "grad_norm": 2.018761157989502,
+      "learning_rate": 0.00016572935801422448,
+      "loss": 2.1472,
+      "step": 7280
+    },
+    {
+      "epoch": 0.17168291649003814,
+      "grad_norm": 1.9133548736572266,
+      "learning_rate": 0.00016568225707691586,
+      "loss": 2.1242,
+      "step": 7290
+    },
+    {
+      "epoch": 0.1719184211765814,
+      "grad_norm": 4.398839473724365,
+      "learning_rate": 0.00016563515613960718,
+      "loss": 2.3578,
+      "step": 7300
+    },
+    {
+      "epoch": 0.17215392586312467,
+      "grad_norm": 1.720661997795105,
+      "learning_rate": 0.00016558805520229853,
+      "loss": 2.2148,
+      "step": 7310
+    },
+    {
+      "epoch": 0.17238943054966793,
+      "grad_norm": 2.4706451892852783,
+      "learning_rate": 0.00016554095426498988,
+      "loss": 2.0974,
+      "step": 7320
+    },
+    {
+      "epoch": 0.1726249352362112,
+      "grad_norm": 2.2956461906433105,
+      "learning_rate": 0.00016549385332768123,
+      "loss": 2.1426,
+      "step": 7330
+    },
+    {
+      "epoch": 0.17286043992275446,
+      "grad_norm": 2.160515069961548,
+      "learning_rate": 0.00016544675239037258,
+      "loss": 2.1641,
+      "step": 7340
+    },
+    {
+      "epoch": 0.17309594460929772,
+      "grad_norm": 2.1589348316192627,
+      "learning_rate": 0.00016539965145306393,
+      "loss": 2.2921,
+      "step": 7350
+    },
+    {
+      "epoch": 0.173331449295841,
+      "grad_norm": 2.136258125305176,
+      "learning_rate": 0.00016535255051575525,
+      "loss": 2.2571,
+      "step": 7360
+    },
+    {
+      "epoch": 0.17356695398238425,
+      "grad_norm": 3.0886502265930176,
+      "learning_rate": 0.00016530544957844663,
+      "loss": 2.3636,
+      "step": 7370
+    },
+    {
+      "epoch": 0.17380245866892752,
+      "grad_norm": 1.8283052444458008,
+      "learning_rate": 0.00016525834864113795,
+      "loss": 2.2302,
+      "step": 7380
+    },
+    {
+      "epoch": 0.17403796335547078,
+      "grad_norm": 2.7342529296875,
+      "learning_rate": 0.00016521124770382933,
+      "loss": 2.32,
+      "step": 7390
+    },
+    {
+      "epoch": 0.17427346804201405,
+      "grad_norm": 2.8123838901519775,
+      "learning_rate": 0.00016516414676652066,
+      "loss": 2.1852,
+      "step": 7400
+    },
+    {
+      "epoch": 0.1745089727285573,
+      "grad_norm": 1.7874890565872192,
+      "learning_rate": 0.000165117045829212,
+      "loss": 2.1981,
+      "step": 7410
+    },
+    {
+      "epoch": 0.17474447741510057,
+      "grad_norm": 2.6356823444366455,
+      "learning_rate": 0.00016506994489190336,
+      "loss": 2.4149,
+      "step": 7420
+    },
+    {
+      "epoch": 0.1749799821016438,
+      "grad_norm": 2.738905906677246,
+      "learning_rate": 0.00016502755404832557,
+      "loss": 2.0254,
+      "step": 7430
+    },
+    {
+      "epoch": 0.17521548678818707,
+      "grad_norm": 1.7247503995895386,
+      "learning_rate": 0.0001649804531110169,
+      "loss": 2.2169,
+      "step": 7440
+    },
+    {
+      "epoch": 0.17545099147473034,
+      "grad_norm": 2.0531630516052246,
+      "learning_rate": 0.00016493335217370827,
+      "loss": 2.2372,
+      "step": 7450
+    },
+    {
+      "epoch": 0.1756864961612736,
+      "grad_norm": 2.0215067863464355,
+      "learning_rate": 0.0001648862512363996,
+      "loss": 1.8726,
+      "step": 7460
+    },
+    {
+      "epoch": 0.17592200084781687,
+      "grad_norm": 2.5239405632019043,
+      "learning_rate": 0.00016483915029909095,
+      "loss": 1.8519,
+      "step": 7470
+    },
+    {
+      "epoch": 0.17615750553436013,
+      "grad_norm": 2.1495165824890137,
+      "learning_rate": 0.0001647920493617823,
+      "loss": 2.1613,
+      "step": 7480
+    },
+    {
+      "epoch": 0.1763930102209034,
+      "grad_norm": 1.7801376581192017,
+      "learning_rate": 0.00016474494842447365,
+      "loss": 2.0669,
+      "step": 7490
+    },
+    {
+      "epoch": 0.17662851490744666,
+      "grad_norm": 2.3613669872283936,
+      "learning_rate": 0.00016469784748716503,
+      "loss": 2.1535,
+      "step": 7500
+    },
+    {
+      "epoch": 0.17686401959398992,
+      "grad_norm": 1.718843936920166,
+      "learning_rate": 0.00016465074654985635,
+      "loss": 2.0689,
+      "step": 7510
+    },
+    {
+      "epoch": 0.1770995242805332,
+      "grad_norm": 1.8729385137557983,
+      "learning_rate": 0.0001646036456125477,
+      "loss": 2.0943,
+      "step": 7520
+    },
+    {
+      "epoch": 0.17733502896707645,
+      "grad_norm": 2.1967837810516357,
+      "learning_rate": 0.00016455654467523905,
+      "loss": 1.9327,
+      "step": 7530
+    },
+    {
+      "epoch": 0.17757053365361972,
+      "grad_norm": 2.0140233039855957,
+      "learning_rate": 0.0001645094437379304,
+      "loss": 2.1792,
+      "step": 7540
+    },
+    {
+      "epoch": 0.17780603834016298,
+      "grad_norm": 1.8183659315109253,
+      "learning_rate": 0.00016446234280062175,
+      "loss": 2.1262,
+      "step": 7550
+    },
+    {
+      "epoch": 0.17804154302670624,
+      "grad_norm": 2.026036500930786,
+      "learning_rate": 0.0001644152418633131,
+      "loss": 1.9423,
+      "step": 7560
+    },
+    {
+      "epoch": 0.17827704771324948,
+      "grad_norm": 3.2327709197998047,
+      "learning_rate": 0.00016436814092600442,
+      "loss": 2.1003,
+      "step": 7570
+    },
+    {
+      "epoch": 0.17851255239979275,
+      "grad_norm": 2.2476227283477783,
+      "learning_rate": 0.0001643210399886958,
+      "loss": 2.1822,
+      "step": 7580
+    },
+    {
+      "epoch": 0.178748057086336,
+      "grad_norm": 2.49985933303833,
+      "learning_rate": 0.00016427393905138712,
+      "loss": 2.1069,
+      "step": 7590
+    },
+    {
+      "epoch": 0.17898356177287927,
+      "grad_norm": 1.9173133373260498,
+      "learning_rate": 0.00016422683811407847,
+      "loss": 2.2727,
+      "step": 7600
+    },
+    {
+      "epoch": 0.17921906645942254,
+      "grad_norm": 2.5073535442352295,
+      "learning_rate": 0.00016417973717676982,
+      "loss": 2.1479,
+      "step": 7610
+    },
+    {
+      "epoch": 0.1794545711459658,
+      "grad_norm": 1.6549034118652344,
+      "learning_rate": 0.00016413263623946117,
+      "loss": 2.1405,
+      "step": 7620
+    },
+    {
+      "epoch": 0.17969007583250907,
+      "grad_norm": 2.1289401054382324,
+      "learning_rate": 0.00016408553530215252,
+      "loss": 2.1708,
+      "step": 7630
+    },
+    {
+      "epoch": 0.17992558051905233,
+      "grad_norm": 2.2387378215789795,
+      "learning_rate": 0.00016403843436484387,
+      "loss": 2.0243,
+      "step": 7640
+    },
+    {
+      "epoch": 0.1801610852055956,
+      "grad_norm": 1.905182123184204,
+      "learning_rate": 0.0001639913334275352,
+      "loss": 2.2722,
+      "step": 7650
+    },
+    {
+      "epoch": 0.18039658989213886,
+      "grad_norm": 2.26442289352417,
+      "learning_rate": 0.00016394423249022658,
+      "loss": 2.1395,
+      "step": 7660
+    },
+    {
+      "epoch": 0.18063209457868212,
+      "grad_norm": 2.4635813236236572,
+      "learning_rate": 0.0001638971315529179,
+      "loss": 2.0492,
+      "step": 7670
+    },
+    {
+      "epoch": 0.1808675992652254,
+      "grad_norm": 2.116964101791382,
+      "learning_rate": 0.00016385003061560928,
+      "loss": 2.1899,
+      "step": 7680
+    },
+    {
+      "epoch": 0.18110310395176865,
+      "grad_norm": 2.431391477584839,
+      "learning_rate": 0.0001638029296783006,
+      "loss": 2.0025,
+      "step": 7690
+    },
+    {
+      "epoch": 0.18133860863831192,
+      "grad_norm": 2.341212749481201,
+      "learning_rate": 0.00016375582874099195,
+      "loss": 2.1425,
+      "step": 7700
+    },
+    {
+      "epoch": 0.18157411332485515,
+      "grad_norm": 1.8115952014923096,
+      "learning_rate": 0.0001637087278036833,
+      "loss": 2.038,
+      "step": 7710
+    },
+    {
+      "epoch": 0.18180961801139842,
+      "grad_norm": 1.9707318544387817,
+      "learning_rate": 0.00016366162686637465,
+      "loss": 2.1576,
+      "step": 7720
+    },
+    {
+      "epoch": 0.18204512269794168,
+      "grad_norm": 2.6205427646636963,
+      "learning_rate": 0.000163614525929066,
+      "loss": 2.2071,
+      "step": 7730
+    },
+    {
+      "epoch": 0.18228062738448494,
+      "grad_norm": 2.613771677017212,
+      "learning_rate": 0.00016356742499175735,
+      "loss": 2.2306,
+      "step": 7740
+    },
+    {
+      "epoch": 0.1825161320710282,
+      "grad_norm": 1.9211950302124023,
+      "learning_rate": 0.00016352032405444867,
+      "loss": 2.1495,
+      "step": 7750
+    },
+    {
+      "epoch": 0.18275163675757147,
+      "grad_norm": 2.385100841522217,
+      "learning_rate": 0.00016347322311714005,
+      "loss": 2.1676,
+      "step": 7760
+    },
+    {
+      "epoch": 0.18298714144411474,
+      "grad_norm": 2.1002583503723145,
+      "learning_rate": 0.00016342612217983137,
+      "loss": 2.2365,
+      "step": 7770
+    },
+    {
+      "epoch": 0.183222646130658,
+      "grad_norm": 2.102348566055298,
+      "learning_rate": 0.00016337902124252272,
+      "loss": 2.1476,
+      "step": 7780
+    },
+    {
+      "epoch": 0.18345815081720127,
+      "grad_norm": 2.3299968242645264,
+      "learning_rate": 0.0001633319203052141,
+      "loss": 2.1953,
+      "step": 7790
+    },
+    {
+      "epoch": 0.18369365550374453,
+      "grad_norm": 1.9959839582443237,
+      "learning_rate": 0.00016328481936790542,
+      "loss": 2.0821,
+      "step": 7800
+    },
+    {
+      "epoch": 0.1839291601902878,
+      "grad_norm": 2.4026405811309814,
+      "learning_rate": 0.00016323771843059677,
+      "loss": 2.1747,
+      "step": 7810
+    },
+    {
+      "epoch": 0.18416466487683106,
+      "grad_norm": 2.223411798477173,
+      "learning_rate": 0.00016319061749328812,
+      "loss": 2.0871,
+      "step": 7820
+    },
+    {
+      "epoch": 0.18440016956337432,
+      "grad_norm": 2.1610796451568604,
+      "learning_rate": 0.00016314351655597948,
+      "loss": 2.1532,
+      "step": 7830
+    },
+    {
+      "epoch": 0.1846356742499176,
+      "grad_norm": 2.226919174194336,
+      "learning_rate": 0.00016309641561867083,
+      "loss": 2.2085,
+      "step": 7840
+    },
+    {
+      "epoch": 0.18487117893646082,
+      "grad_norm": 2.2085070610046387,
+      "learning_rate": 0.00016304931468136218,
+      "loss": 2.2106,
+      "step": 7850
+    },
+    {
+      "epoch": 0.1851066836230041,
+      "grad_norm": 2.343294858932495,
+      "learning_rate": 0.0001630022137440535,
+      "loss": 2.128,
+      "step": 7860
+    },
+    {
+      "epoch": 0.18534218830954735,
+      "grad_norm": 1.8624409437179565,
+      "learning_rate": 0.00016295511280674488,
+      "loss": 2.1655,
+      "step": 7870
+    },
+    {
+      "epoch": 0.18557769299609062,
+      "grad_norm": 2.0223777294158936,
+      "learning_rate": 0.0001629080118694362,
+      "loss": 2.1523,
+      "step": 7880
+    },
+    {
+      "epoch": 0.18581319768263388,
+      "grad_norm": 2.1489696502685547,
+      "learning_rate": 0.00016286091093212758,
+      "loss": 2.1188,
+      "step": 7890
+    },
+    {
+      "epoch": 0.18604870236917714,
+      "grad_norm": 2.2076354026794434,
+      "learning_rate": 0.0001628138099948189,
+      "loss": 1.9737,
+      "step": 7900
+    },
+    {
+      "epoch": 0.1862842070557204,
+      "grad_norm": 2.566239595413208,
+      "learning_rate": 0.00016276670905751025,
+      "loss": 2.2254,
+      "step": 7910
+    },
+    {
+      "epoch": 0.18651971174226367,
+      "grad_norm": 2.0568361282348633,
+      "learning_rate": 0.0001627196081202016,
+      "loss": 2.1747,
+      "step": 7920
+    },
+    {
+      "epoch": 0.18675521642880694,
+      "grad_norm": 2.4507033824920654,
+      "learning_rate": 0.00016267250718289295,
+      "loss": 2.3182,
+      "step": 7930
+    },
+    {
+      "epoch": 0.1869907211153502,
+      "grad_norm": 2.1586856842041016,
+      "learning_rate": 0.0001626254062455843,
+      "loss": 2.1965,
+      "step": 7940
+    },
+    {
+      "epoch": 0.18722622580189346,
+      "grad_norm": 1.9273881912231445,
+      "learning_rate": 0.00016257830530827565,
+      "loss": 2.2813,
+      "step": 7950
+    },
+    {
+      "epoch": 0.18746173048843673,
+      "grad_norm": 1.7580167055130005,
+      "learning_rate": 0.00016253120437096697,
+      "loss": 2.1578,
+      "step": 7960
+    },
+    {
+      "epoch": 0.18769723517498,
+      "grad_norm": 2.0554792881011963,
+      "learning_rate": 0.00016248410343365835,
+      "loss": 2.0389,
+      "step": 7970
+    },
+    {
+      "epoch": 0.18793273986152326,
+      "grad_norm": 1.872406005859375,
+      "learning_rate": 0.00016243700249634967,
+      "loss": 2.1407,
+      "step": 7980
+    },
+    {
+      "epoch": 0.1881682445480665,
+      "grad_norm": 2.1764724254608154,
+      "learning_rate": 0.00016238990155904102,
+      "loss": 2.135,
+      "step": 7990
+    },
+    {
+      "epoch": 0.18840374923460976,
+      "grad_norm": 1.8471111059188843,
+      "learning_rate": 0.00016234280062173237,
+      "loss": 2.1204,
+      "step": 8000
+    },
+    {
+      "epoch": 0.18863925392115302,
+      "grad_norm": 2.7460036277770996,
+      "learning_rate": 0.00016229569968442373,
+      "loss": 2.1958,
+      "step": 8010
+    },
+    {
+      "epoch": 0.18887475860769629,
+      "grad_norm": 1.8986886739730835,
+      "learning_rate": 0.00016224859874711508,
+      "loss": 2.2771,
+      "step": 8020
+    },
+    {
+      "epoch": 0.18911026329423955,
+      "grad_norm": 2.3261477947235107,
+      "learning_rate": 0.00016220149780980643,
+      "loss": 2.2911,
+      "step": 8030
+    },
+    {
+      "epoch": 0.18934576798078281,
+      "grad_norm": 3.4282009601593018,
+      "learning_rate": 0.00016215439687249775,
+      "loss": 2.1043,
+      "step": 8040
+    },
+    {
+      "epoch": 0.18958127266732608,
+      "grad_norm": 1.8405457735061646,
+      "learning_rate": 0.00016210729593518913,
+      "loss": 2.211,
+      "step": 8050
+    },
+    {
+      "epoch": 0.18981677735386934,
+      "grad_norm": 2.0625762939453125,
+      "learning_rate": 0.00016206019499788045,
+      "loss": 2.0193,
+      "step": 8060
+    },
+    {
+      "epoch": 0.1900522820404126,
+      "grad_norm": 1.9741464853286743,
+      "learning_rate": 0.0001620130940605718,
+      "loss": 2.1448,
+      "step": 8070
+    },
+    {
+      "epoch": 0.19028778672695587,
+      "grad_norm": 2.270921468734741,
+      "learning_rate": 0.00016196599312326318,
+      "loss": 2.3037,
+      "step": 8080
+    },
+    {
+      "epoch": 0.19052329141349914,
+      "grad_norm": 1.8715896606445312,
+      "learning_rate": 0.0001619188921859545,
+      "loss": 2.027,
+      "step": 8090
+    },
+    {
+      "epoch": 0.1907587961000424,
+      "grad_norm": 2.515016794204712,
+      "learning_rate": 0.00016187179124864588,
+      "loss": 1.9652,
+      "step": 8100
+    },
+    {
+      "epoch": 0.19099430078658566,
+      "grad_norm": 1.9395673274993896,
+      "learning_rate": 0.0001618246903113372,
+      "loss": 2.2969,
+      "step": 8110
+    },
+    {
+      "epoch": 0.19122980547312893,
+      "grad_norm": 2.1764769554138184,
+      "learning_rate": 0.00016177758937402855,
+      "loss": 2.1232,
+      "step": 8120
+    },
+    {
+      "epoch": 0.19146531015967216,
+      "grad_norm": 2.594571113586426,
+      "learning_rate": 0.0001617304884367199,
+      "loss": 2.1088,
+      "step": 8130
+    },
+    {
+      "epoch": 0.19170081484621543,
+      "grad_norm": 2.394315242767334,
+      "learning_rate": 0.00016168338749941125,
+      "loss": 2.0626,
+      "step": 8140
+    },
+    {
+      "epoch": 0.1919363195327587,
+      "grad_norm": 1.8806692361831665,
+      "learning_rate": 0.0001616362865621026,
+      "loss": 2.0899,
+      "step": 8150
+    },
+    {
+      "epoch": 0.19217182421930196,
+      "grad_norm": 2.2161996364593506,
+      "learning_rate": 0.00016158918562479395,
+      "loss": 2.2571,
+      "step": 8160
+    },
+    {
+      "epoch": 0.19240732890584522,
+      "grad_norm": 3.4953277111053467,
+      "learning_rate": 0.00016154208468748527,
+      "loss": 2.0105,
+      "step": 8170
+    },
+    {
+      "epoch": 0.19264283359238848,
+      "grad_norm": 2.8051047325134277,
+      "learning_rate": 0.00016149498375017665,
+      "loss": 2.1364,
+      "step": 8180
+    },
+    {
+      "epoch": 0.19287833827893175,
+      "grad_norm": 2.5728232860565186,
+      "learning_rate": 0.00016144788281286798,
+      "loss": 2.0509,
+      "step": 8190
+    },
+    {
+      "epoch": 0.193113842965475,
+      "grad_norm": 2.346428394317627,
+      "learning_rate": 0.00016140078187555933,
+      "loss": 2.2152,
+      "step": 8200
+    },
+    {
+      "epoch": 0.19334934765201828,
+      "grad_norm": 1.9895967245101929,
+      "learning_rate": 0.00016135368093825068,
+      "loss": 2.1382,
+      "step": 8210
+    },
+    {
+      "epoch": 0.19358485233856154,
+      "grad_norm": 2.0366828441619873,
+      "learning_rate": 0.00016130658000094203,
+      "loss": 2.1697,
+      "step": 8220
+    },
+    {
+      "epoch": 0.1938203570251048,
+      "grad_norm": 2.5013720989227295,
+      "learning_rate": 0.00016125947906363338,
+      "loss": 1.9348,
+      "step": 8230
+    },
+    {
+      "epoch": 0.19405586171164807,
+      "grad_norm": 2.338641881942749,
+      "learning_rate": 0.00016121237812632473,
+      "loss": 2.3867,
+      "step": 8240
+    },
+    {
+      "epoch": 0.19429136639819133,
+      "grad_norm": 2.367546319961548,
+      "learning_rate": 0.00016116527718901605,
+      "loss": 2.1588,
+      "step": 8250
+    },
+    {
+      "epoch": 0.1945268710847346,
+      "grad_norm": 2.366485118865967,
+      "learning_rate": 0.00016111817625170743,
+      "loss": 2.3467,
+      "step": 8260
+    },
+    {
+      "epoch": 0.19476237577127783,
+      "grad_norm": 1.8073631525039673,
+      "learning_rate": 0.00016107107531439875,
+      "loss": 2.0071,
+      "step": 8270
+    },
+    {
+      "epoch": 0.1949978804578211,
+      "grad_norm": 1.8550753593444824,
+      "learning_rate": 0.00016102397437709013,
+      "loss": 2.1192,
+      "step": 8280
+    },
+    {
+      "epoch": 0.19523338514436436,
+      "grad_norm": 2.834643602371216,
+      "learning_rate": 0.00016097687343978145,
+      "loss": 2.1614,
+      "step": 8290
+    },
+    {
+      "epoch": 0.19546888983090763,
+      "grad_norm": 1.961897850036621,
+      "learning_rate": 0.0001609297725024728,
+      "loss": 2.1747,
+      "step": 8300
+    },
+    {
+      "epoch": 0.1957043945174509,
+      "grad_norm": 2.972184658050537,
+      "learning_rate": 0.00016088267156516415,
+      "loss": 2.1692,
+      "step": 8310
+    },
+    {
+      "epoch": 0.19593989920399416,
+      "grad_norm": 2.2215375900268555,
+      "learning_rate": 0.0001608355706278555,
+      "loss": 2.1288,
+      "step": 8320
+    },
+    {
+      "epoch": 0.19617540389053742,
+      "grad_norm": 1.6455684900283813,
+      "learning_rate": 0.00016078846969054685,
+      "loss": 2.0821,
+      "step": 8330
+    },
+    {
+      "epoch": 0.19641090857708068,
+      "grad_norm": 3.122081995010376,
+      "learning_rate": 0.0001607413687532382,
+      "loss": 2.2687,
+      "step": 8340
+    },
+    {
+      "epoch": 0.19664641326362395,
+      "grad_norm": 2.8040692806243896,
+      "learning_rate": 0.00016069426781592955,
+      "loss": 2.3317,
+      "step": 8350
+    },
+    {
+      "epoch": 0.1968819179501672,
+      "grad_norm": 2.200100898742676,
+      "learning_rate": 0.0001606471668786209,
+      "loss": 2.1458,
+      "step": 8360
+    },
+    {
+      "epoch": 0.19711742263671048,
+      "grad_norm": 1.9488954544067383,
+      "learning_rate": 0.00016060006594131225,
+      "loss": 1.9935,
+      "step": 8370
+    },
+    {
+      "epoch": 0.19735292732325374,
+      "grad_norm": 2.2211225032806396,
+      "learning_rate": 0.00016055296500400358,
+      "loss": 2.3222,
+      "step": 8380
+    },
+    {
+      "epoch": 0.197588432009797,
+      "grad_norm": 2.1487669944763184,
+      "learning_rate": 0.00016050586406669495,
+      "loss": 2.1131,
+      "step": 8390
+    },
+    {
+      "epoch": 0.19782393669634027,
+      "grad_norm": 2.1569080352783203,
+      "learning_rate": 0.00016045876312938628,
+      "loss": 2.184,
+      "step": 8400
+    },
+    {
+      "epoch": 0.19805944138288353,
+      "grad_norm": 1.7691841125488281,
+      "learning_rate": 0.00016041166219207763,
+      "loss": 2.2211,
+      "step": 8410
+    },
+    {
+      "epoch": 0.19829494606942677,
+      "grad_norm": 2.2615177631378174,
+      "learning_rate": 0.00016036456125476898,
+      "loss": 2.171,
+      "step": 8420
+    },
+    {
+      "epoch": 0.19853045075597003,
+      "grad_norm": 1.9963940382003784,
+      "learning_rate": 0.00016031746031746033,
+      "loss": 2.0384,
+      "step": 8430
+    },
+    {
+      "epoch": 0.1987659554425133,
+      "grad_norm": 2.1423399448394775,
+      "learning_rate": 0.00016027035938015168,
+      "loss": 2.1735,
+      "step": 8440
+    },
+    {
+      "epoch": 0.19900146012905656,
+      "grad_norm": 2.056474447250366,
+      "learning_rate": 0.00016022325844284303,
+      "loss": 2.2922,
+      "step": 8450
+    },
+    {
+      "epoch": 0.19923696481559983,
+      "grad_norm": 1.6666935682296753,
+      "learning_rate": 0.00016017615750553435,
+      "loss": 2.1775,
+      "step": 8460
+    },
+    {
+      "epoch": 0.1994724695021431,
+      "grad_norm": 2.2236125469207764,
+      "learning_rate": 0.00016012905656822573,
+      "loss": 1.9681,
+      "step": 8470
+    },
+    {
+      "epoch": 0.19970797418868635,
+      "grad_norm": 2.01289701461792,
+      "learning_rate": 0.00016008195563091705,
+      "loss": 1.9393,
+      "step": 8480
+    },
+    {
+      "epoch": 0.19994347887522962,
+      "grad_norm": 2.205111503601074,
+      "learning_rate": 0.00016003485469360843,
+      "loss": 1.9858,
+      "step": 8490
+    },
+    {
+      "epoch": 0.20017898356177288,
+      "grad_norm": 3.152418375015259,
+      "learning_rate": 0.00015998775375629975,
+      "loss": 2.241,
+      "step": 8500
+    },
+    {
+      "epoch": 0.20041448824831615,
+      "grad_norm": 2.2211573123931885,
+      "learning_rate": 0.0001599406528189911,
+      "loss": 2.0439,
+      "step": 8510
+    },
+    {
+      "epoch": 0.2006499929348594,
+      "grad_norm": 1.9449644088745117,
+      "learning_rate": 0.00015989355188168245,
+      "loss": 2.0677,
+      "step": 8520
+    },
+    {
+      "epoch": 0.20088549762140268,
+      "grad_norm": 3.1177256107330322,
+      "learning_rate": 0.0001598464509443738,
+      "loss": 2.1057,
+      "step": 8530
+    },
+    {
+      "epoch": 0.20112100230794594,
+      "grad_norm": 3.570312023162842,
+      "learning_rate": 0.00015979935000706515,
+      "loss": 2.0658,
+      "step": 8540
+    },
+    {
+      "epoch": 0.2013565069944892,
+      "grad_norm": 1.7642455101013184,
+      "learning_rate": 0.0001597522490697565,
+      "loss": 2.2294,
+      "step": 8550
+    },
+    {
+      "epoch": 0.20159201168103244,
+      "grad_norm": 1.7832077741622925,
+      "learning_rate": 0.00015970514813244783,
+      "loss": 2.2795,
+      "step": 8560
+    },
+    {
+      "epoch": 0.2018275163675757,
+      "grad_norm": 2.377683639526367,
+      "learning_rate": 0.0001596580471951392,
+      "loss": 2.0018,
+      "step": 8570
+    },
+    {
+      "epoch": 0.20206302105411897,
+      "grad_norm": 2.7530715465545654,
+      "learning_rate": 0.00015961094625783053,
+      "loss": 2.2075,
+      "step": 8580
+    },
+    {
+      "epoch": 0.20229852574066223,
+      "grad_norm": 2.924802303314209,
+      "learning_rate": 0.00015956384532052188,
+      "loss": 2.2061,
+      "step": 8590
+    },
+    {
+      "epoch": 0.2025340304272055,
+      "grad_norm": 2.2290501594543457,
+      "learning_rate": 0.00015951674438321323,
+      "loss": 1.9976,
+      "step": 8600
+    },
+    {
+      "epoch": 0.20276953511374876,
+      "grad_norm": 2.3307456970214844,
+      "learning_rate": 0.00015946964344590458,
+      "loss": 2.1666,
+      "step": 8610
+    },
+    {
+      "epoch": 0.20300503980029203,
+      "grad_norm": 2.272770643234253,
+      "learning_rate": 0.00015942254250859593,
+      "loss": 2.1245,
+      "step": 8620
+    },
+    {
+      "epoch": 0.2032405444868353,
+      "grad_norm": 2.5355868339538574,
+      "learning_rate": 0.00015937544157128728,
+      "loss": 2.2273,
+      "step": 8630
+    },
+    {
+      "epoch": 0.20347604917337855,
+      "grad_norm": 1.9374130964279175,
+      "learning_rate": 0.00015932834063397863,
+      "loss": 1.9849,
+      "step": 8640
+    },
+    {
+      "epoch": 0.20371155385992182,
+      "grad_norm": 2.0715346336364746,
+      "learning_rate": 0.00015928123969666998,
+      "loss": 1.9397,
+      "step": 8650
+    },
+    {
+      "epoch": 0.20394705854646508,
+      "grad_norm": 2.3634729385375977,
+      "learning_rate": 0.00015923413875936133,
+      "loss": 2.1056,
+      "step": 8660
+    },
+    {
+      "epoch": 0.20418256323300835,
+      "grad_norm": 2.003783702850342,
+      "learning_rate": 0.00015918703782205265,
+      "loss": 1.8101,
+      "step": 8670
+    },
+    {
+      "epoch": 0.2044180679195516,
+      "grad_norm": 1.9855163097381592,
+      "learning_rate": 0.00015913993688474403,
+      "loss": 1.8513,
+      "step": 8680
+    },
+    {
+      "epoch": 0.20465357260609487,
+      "grad_norm": 2.3215885162353516,
+      "learning_rate": 0.00015909283594743535,
+      "loss": 2.2075,
+      "step": 8690
+    },
+    {
+      "epoch": 0.2048890772926381,
+      "grad_norm": 2.110724687576294,
+      "learning_rate": 0.00015904573501012673,
+      "loss": 2.0604,
+      "step": 8700
+    },
+    {
+      "epoch": 0.20512458197918138,
+      "grad_norm": 2.33134388923645,
+      "learning_rate": 0.00015899863407281805,
+      "loss": 2.1537,
+      "step": 8710
+    },
+    {
+      "epoch": 0.20536008666572464,
+      "grad_norm": 2.0391716957092285,
+      "learning_rate": 0.0001589515331355094,
+      "loss": 1.9752,
+      "step": 8720
+    },
+    {
+      "epoch": 0.2055955913522679,
+      "grad_norm": 2.1670148372650146,
+      "learning_rate": 0.00015890443219820075,
+      "loss": 2.1016,
+      "step": 8730
+    },
+    {
+      "epoch": 0.20583109603881117,
+      "grad_norm": 2.2209887504577637,
+      "learning_rate": 0.0001588573312608921,
+      "loss": 2.1446,
+      "step": 8740
+    },
+    {
+      "epoch": 0.20606660072535443,
+      "grad_norm": 2.245386838912964,
+      "learning_rate": 0.00015881023032358345,
+      "loss": 1.9711,
+      "step": 8750
+    },
+    {
+      "epoch": 0.2063021054118977,
+      "grad_norm": 2.2087981700897217,
+      "learning_rate": 0.0001587631293862748,
+      "loss": 2.157,
+      "step": 8760
+    },
+    {
+      "epoch": 0.20653761009844096,
+      "grad_norm": 1.9615005254745483,
+      "learning_rate": 0.00015871602844896613,
+      "loss": 2.0809,
+      "step": 8770
+    },
+    {
+      "epoch": 0.20677311478498422,
+      "grad_norm": 2.1941707134246826,
+      "learning_rate": 0.0001586689275116575,
+      "loss": 2.0107,
+      "step": 8780
+    },
+    {
+      "epoch": 0.2070086194715275,
+      "grad_norm": 1.8866039514541626,
+      "learning_rate": 0.00015862182657434883,
+      "loss": 2.0821,
+      "step": 8790
+    },
+    {
+      "epoch": 0.20724412415807075,
+      "grad_norm": 2.4562177658081055,
+      "learning_rate": 0.00015857472563704018,
+      "loss": 2.2769,
+      "step": 8800
+    },
+    {
+      "epoch": 0.20747962884461402,
+      "grad_norm": 1.8872411251068115,
+      "learning_rate": 0.00015852762469973153,
+      "loss": 2.307,
+      "step": 8810
+    },
+    {
+      "epoch": 0.20771513353115728,
+      "grad_norm": 1.9673904180526733,
+      "learning_rate": 0.00015848052376242288,
+      "loss": 2.242,
+      "step": 8820
+    },
+    {
+      "epoch": 0.20795063821770055,
+      "grad_norm": 2.0321238040924072,
+      "learning_rate": 0.00015843342282511423,
+      "loss": 2.0609,
+      "step": 8830
+    },
+    {
+      "epoch": 0.20818614290424378,
+      "grad_norm": 2.2742855548858643,
+      "learning_rate": 0.00015838632188780558,
+      "loss": 2.3143,
+      "step": 8840
+    },
+    {
+      "epoch": 0.20842164759078705,
+      "grad_norm": 2.290383815765381,
+      "learning_rate": 0.0001583392209504969,
+      "loss": 2.2528,
+      "step": 8850
+    },
+    {
+      "epoch": 0.2086571522773303,
+      "grad_norm": 2.4600648880004883,
+      "learning_rate": 0.00015829212001318828,
+      "loss": 2.3136,
+      "step": 8860
+    },
+    {
+      "epoch": 0.20889265696387357,
+      "grad_norm": 2.9521021842956543,
+      "learning_rate": 0.0001582450190758796,
+      "loss": 2.3804,
+      "step": 8870
+    },
+    {
+      "epoch": 0.20912816165041684,
+      "grad_norm": 2.5371804237365723,
+      "learning_rate": 0.00015819791813857098,
+      "loss": 2.008,
+      "step": 8880
+    },
+    {
+      "epoch": 0.2093636663369601,
+      "grad_norm": 1.9548083543777466,
+      "learning_rate": 0.0001581508172012623,
+      "loss": 2.2387,
+      "step": 8890
+    },
+    {
+      "epoch": 0.20959917102350337,
+      "grad_norm": 2.33406925201416,
+      "learning_rate": 0.00015810371626395365,
+      "loss": 1.9671,
+      "step": 8900
+    },
+    {
+      "epoch": 0.20983467571004663,
+      "grad_norm": 2.2755494117736816,
+      "learning_rate": 0.000158056615326645,
+      "loss": 2.1114,
+      "step": 8910
+    },
+    {
+      "epoch": 0.2100701803965899,
+      "grad_norm": 2.0145130157470703,
+      "learning_rate": 0.00015800951438933635,
+      "loss": 2.1826,
+      "step": 8920
+    },
+    {
+      "epoch": 0.21030568508313316,
+      "grad_norm": 2.4278318881988525,
+      "learning_rate": 0.0001579624134520277,
+      "loss": 2.2274,
+      "step": 8930
+    },
+    {
+      "epoch": 0.21054118976967642,
+      "grad_norm": 2.743621826171875,
+      "learning_rate": 0.00015791531251471905,
+      "loss": 2.03,
+      "step": 8940
+    },
+    {
+      "epoch": 0.2107766944562197,
+      "grad_norm": 2.0403552055358887,
+      "learning_rate": 0.0001578682115774104,
+      "loss": 1.9389,
+      "step": 8950
+    },
+    {
+      "epoch": 0.21101219914276295,
+      "grad_norm": 1.9980547428131104,
+      "learning_rate": 0.00015782111064010175,
+      "loss": 2.207,
+      "step": 8960
+    },
+    {
+      "epoch": 0.21124770382930622,
+      "grad_norm": 2.2395660877227783,
+      "learning_rate": 0.0001577740097027931,
+      "loss": 2.0992,
+      "step": 8970
+    },
+    {
+      "epoch": 0.21148320851584945,
+      "grad_norm": 2.130779266357422,
+      "learning_rate": 0.00015772690876548443,
+      "loss": 2.1155,
+      "step": 8980
+    },
+    {
+      "epoch": 0.21171871320239272,
+      "grad_norm": 2.452732801437378,
+      "learning_rate": 0.0001576798078281758,
+      "loss": 2.1982,
+      "step": 8990
+    },
+    {
+      "epoch": 0.21195421788893598,
+      "grad_norm": 1.9832158088684082,
+      "learning_rate": 0.00015763270689086713,
+      "loss": 2.0777,
+      "step": 9000
+    },
+    {
+      "epoch": 0.21218972257547924,
+      "grad_norm": 2.609037160873413,
+      "learning_rate": 0.00015758560595355848,
+      "loss": 2.2989,
+      "step": 9010
+    },
+    {
+      "epoch": 0.2124252272620225,
+      "grad_norm": 2.5825107097625732,
+      "learning_rate": 0.00015753850501624983,
+      "loss": 2.1168,
+      "step": 9020
+    },
+    {
+      "epoch": 0.21266073194856577,
+      "grad_norm": 2.1648106575012207,
+      "learning_rate": 0.00015749140407894118,
+      "loss": 2.1504,
+      "step": 9030
+    },
+    {
+      "epoch": 0.21289623663510904,
+      "grad_norm": 2.132033586502075,
+      "learning_rate": 0.00015744430314163253,
+      "loss": 2.2418,
+      "step": 9040
+    },
+    {
+      "epoch": 0.2131317413216523,
+      "grad_norm": 2.439497470855713,
+      "learning_rate": 0.00015739720220432388,
+      "loss": 2.2041,
+      "step": 9050
+    },
+    {
+      "epoch": 0.21336724600819557,
+      "grad_norm": 2.2189526557922363,
+      "learning_rate": 0.0001573501012670152,
+      "loss": 2.1985,
+      "step": 9060
+    },
+    {
+      "epoch": 0.21360275069473883,
+      "grad_norm": 2.7584781646728516,
+      "learning_rate": 0.00015730300032970658,
+      "loss": 2.2038,
+      "step": 9070
+    },
+    {
+      "epoch": 0.2138382553812821,
+      "grad_norm": 2.2705740928649902,
+      "learning_rate": 0.0001572558993923979,
+      "loss": 2.1067,
+      "step": 9080
+    },
+    {
+      "epoch": 0.21407376006782536,
+      "grad_norm": 3.012639045715332,
+      "learning_rate": 0.00015720879845508928,
+      "loss": 2.3559,
+      "step": 9090
+    },
+    {
+      "epoch": 0.21430926475436862,
+      "grad_norm": 3.75484299659729,
+      "learning_rate": 0.0001571616975177806,
+      "loss": 2.1761,
+      "step": 9100
+    },
+    {
+      "epoch": 0.2145447694409119,
+      "grad_norm": 2.1686441898345947,
+      "learning_rate": 0.00015711459658047195,
+      "loss": 2.1525,
+      "step": 9110
+    },
+    {
+      "epoch": 0.21478027412745512,
+      "grad_norm": 2.0703907012939453,
+      "learning_rate": 0.0001570674956431633,
+      "loss": 2.3426,
+      "step": 9120
+    },
+    {
+      "epoch": 0.2150157788139984,
+      "grad_norm": 2.633394956588745,
+      "learning_rate": 0.00015702039470585465,
+      "loss": 2.101,
+      "step": 9130
+    },
+    {
+      "epoch": 0.21525128350054165,
+      "grad_norm": 1.6758605241775513,
+      "learning_rate": 0.000156973293768546,
+      "loss": 2.422,
+      "step": 9140
+    },
+    {
+      "epoch": 0.21548678818708492,
+      "grad_norm": 2.618814468383789,
+      "learning_rate": 0.00015692619283123735,
+      "loss": 2.2528,
+      "step": 9150
+    },
+    {
+      "epoch": 0.21572229287362818,
+      "grad_norm": 1.8236844539642334,
+      "learning_rate": 0.00015687909189392868,
+      "loss": 2.075,
+      "step": 9160
+    },
+    {
+      "epoch": 0.21595779756017144,
+      "grad_norm": 2.2664949893951416,
+      "learning_rate": 0.00015683199095662006,
+      "loss": 1.9622,
+      "step": 9170
+    },
+    {
+      "epoch": 0.2161933022467147,
+      "grad_norm": 2.9308884143829346,
+      "learning_rate": 0.00015678489001931138,
+      "loss": 2.0478,
+      "step": 9180
+    },
+    {
+      "epoch": 0.21642880693325797,
+      "grad_norm": 2.6162636280059814,
+      "learning_rate": 0.00015673778908200273,
+      "loss": 2.2457,
+      "step": 9190
+    },
+    {
+      "epoch": 0.21666431161980124,
+      "grad_norm": 2.512709617614746,
+      "learning_rate": 0.0001566906881446941,
+      "loss": 2.1442,
+      "step": 9200
+    },
+    {
+      "epoch": 0.2168998163063445,
+      "grad_norm": 1.8639780282974243,
+      "learning_rate": 0.00015664358720738543,
+      "loss": 2.0677,
+      "step": 9210
+    },
+    {
+      "epoch": 0.21713532099288776,
+      "grad_norm": 2.1955885887145996,
+      "learning_rate": 0.00015659648627007678,
+      "loss": 2.3215,
+      "step": 9220
+    },
+    {
+      "epoch": 0.21737082567943103,
+      "grad_norm": 2.1745858192443848,
+      "learning_rate": 0.00015654938533276813,
+      "loss": 2.2667,
+      "step": 9230
+    },
+    {
+      "epoch": 0.2176063303659743,
+      "grad_norm": 2.6070613861083984,
+      "learning_rate": 0.00015650228439545948,
+      "loss": 2.2632,
+      "step": 9240
+    },
+    {
+      "epoch": 0.21784183505251756,
+      "grad_norm": 3.0261740684509277,
+      "learning_rate": 0.00015645518345815083,
+      "loss": 2.0795,
+      "step": 9250
+    },
+    {
+      "epoch": 0.2180773397390608,
+      "grad_norm": 2.801222324371338,
+      "learning_rate": 0.00015640808252084218,
+      "loss": 2.2413,
+      "step": 9260
+    },
+    {
+      "epoch": 0.21831284442560406,
+      "grad_norm": 3.1285176277160645,
+      "learning_rate": 0.0001563609815835335,
+      "loss": 2.2592,
+      "step": 9270
+    },
+    {
+      "epoch": 0.21854834911214732,
+      "grad_norm": 2.106144666671753,
+      "learning_rate": 0.00015631388064622488,
+      "loss": 2.1891,
+      "step": 9280
+    },
+    {
+      "epoch": 0.2187838537986906,
+      "grad_norm": 1.9326168298721313,
+      "learning_rate": 0.0001562667797089162,
+      "loss": 2.0664,
+      "step": 9290
+    },
+    {
+      "epoch": 0.21901935848523385,
+      "grad_norm": 1.709635853767395,
+      "learning_rate": 0.00015621967877160758,
+      "loss": 2.1744,
+      "step": 9300
+    },
+    {
+      "epoch": 0.21925486317177711,
+      "grad_norm": 2.1247639656066895,
+      "learning_rate": 0.0001561725778342989,
+      "loss": 2.1967,
+      "step": 9310
+    },
+    {
+      "epoch": 0.21949036785832038,
+      "grad_norm": 2.582542657852173,
+      "learning_rate": 0.00015612547689699025,
+      "loss": 2.2411,
+      "step": 9320
+    },
+    {
+      "epoch": 0.21972587254486364,
+      "grad_norm": 2.743621349334717,
+      "learning_rate": 0.0001560783759596816,
+      "loss": 2.3035,
+      "step": 9330
+    },
+    {
+      "epoch": 0.2199613772314069,
+      "grad_norm": 2.67402720451355,
+      "learning_rate": 0.00015603127502237296,
+      "loss": 2.2365,
+      "step": 9340
+    },
+    {
+      "epoch": 0.22019688191795017,
+      "grad_norm": 2.4903061389923096,
+      "learning_rate": 0.0001559841740850643,
+      "loss": 2.0761,
+      "step": 9350
+    },
+    {
+      "epoch": 0.22043238660449344,
+      "grad_norm": 2.3421287536621094,
+      "learning_rate": 0.00015593707314775566,
+      "loss": 2.1829,
+      "step": 9360
+    },
+    {
+      "epoch": 0.2206678912910367,
+      "grad_norm": 2.735867977142334,
+      "learning_rate": 0.00015588997221044698,
+      "loss": 2.2231,
+      "step": 9370
+    },
+    {
+      "epoch": 0.22090339597757996,
+      "grad_norm": 1.992004156112671,
+      "learning_rate": 0.00015584287127313836,
+      "loss": 2.2848,
+      "step": 9380
+    },
+    {
+      "epoch": 0.22113890066412323,
+      "grad_norm": 2.300267457962036,
+      "learning_rate": 0.00015579577033582968,
+      "loss": 2.1919,
+      "step": 9390
+    },
+    {
+      "epoch": 0.22137440535066646,
+      "grad_norm": 1.9108951091766357,
+      "learning_rate": 0.00015574866939852103,
+      "loss": 2.0605,
+      "step": 9400
+    },
+    {
+      "epoch": 0.22160991003720973,
+      "grad_norm": 2.468258857727051,
+      "learning_rate": 0.00015570156846121238,
+      "loss": 2.2323,
+      "step": 9410
+    },
+    {
+      "epoch": 0.221845414723753,
+      "grad_norm": 2.3677616119384766,
+      "learning_rate": 0.00015565446752390373,
+      "loss": 2.0305,
+      "step": 9420
+    },
+    {
+      "epoch": 0.22208091941029626,
+      "grad_norm": 2.333322286605835,
+      "learning_rate": 0.00015560736658659508,
+      "loss": 2.3494,
+      "step": 9430
+    },
+    {
+      "epoch": 0.22231642409683952,
+      "grad_norm": 2.4163379669189453,
+      "learning_rate": 0.00015556026564928643,
+      "loss": 2.2143,
+      "step": 9440
+    },
+    {
+      "epoch": 0.22255192878338279,
+      "grad_norm": 2.2694649696350098,
+      "learning_rate": 0.00015551316471197775,
+      "loss": 2.1862,
+      "step": 9450
+    },
+    {
+      "epoch": 0.22278743346992605,
+      "grad_norm": 1.8725945949554443,
+      "learning_rate": 0.00015546606377466913,
+      "loss": 2.1275,
+      "step": 9460
+    },
+    {
+      "epoch": 0.2230229381564693,
+      "grad_norm": 2.3583409786224365,
+      "learning_rate": 0.00015541896283736045,
+      "loss": 2.2181,
+      "step": 9470
+    },
+    {
+      "epoch": 0.22325844284301258,
+      "grad_norm": 1.9677271842956543,
+      "learning_rate": 0.00015537186190005183,
+      "loss": 2.1267,
+      "step": 9480
+    },
+    {
+      "epoch": 0.22349394752955584,
+      "grad_norm": 1.9935779571533203,
+      "learning_rate": 0.00015532476096274318,
+      "loss": 2.3531,
+      "step": 9490
+    },
+    {
+      "epoch": 0.2237294522160991,
+      "grad_norm": 3.433013439178467,
+      "learning_rate": 0.0001552776600254345,
+      "loss": 2.1142,
+      "step": 9500
+    },
+    {
+      "epoch": 0.22396495690264237,
+      "grad_norm": 2.3903839588165283,
+      "learning_rate": 0.00015523055908812588,
+      "loss": 2.0263,
+      "step": 9510
+    },
+    {
+      "epoch": 0.22420046158918563,
+      "grad_norm": 2.319390296936035,
+      "learning_rate": 0.0001551834581508172,
+      "loss": 2.2263,
+      "step": 9520
+    },
+    {
+      "epoch": 0.2244359662757289,
+      "grad_norm": 1.99445641040802,
+      "learning_rate": 0.00015513635721350856,
+      "loss": 1.9634,
+      "step": 9530
+    },
+    {
+      "epoch": 0.22467147096227216,
+      "grad_norm": 2.1183362007141113,
+      "learning_rate": 0.0001550892562761999,
+      "loss": 2.1869,
+      "step": 9540
+    },
+    {
+      "epoch": 0.2249069756488154,
+      "grad_norm": 2.5018179416656494,
+      "learning_rate": 0.00015504215533889126,
+      "loss": 2.0954,
+      "step": 9550
+    },
+    {
+      "epoch": 0.22514248033535866,
+      "grad_norm": 2.222480297088623,
+      "learning_rate": 0.0001549950544015826,
+      "loss": 2.0068,
+      "step": 9560
+    },
+    {
+      "epoch": 0.22537798502190193,
+      "grad_norm": 2.6220204830169678,
+      "learning_rate": 0.00015494795346427396,
+      "loss": 2.19,
+      "step": 9570
+    },
+    {
+      "epoch": 0.2256134897084452,
+      "grad_norm": 2.385401964187622,
+      "learning_rate": 0.00015490085252696528,
+      "loss": 2.1428,
+      "step": 9580
+    },
+    {
+      "epoch": 0.22584899439498846,
+      "grad_norm": 2.0146987438201904,
+      "learning_rate": 0.00015485375158965666,
+      "loss": 2.1736,
+      "step": 9590
+    },
+    {
+      "epoch": 0.22608449908153172,
+      "grad_norm": 2.207779884338379,
+      "learning_rate": 0.00015480665065234798,
+      "loss": 2.0437,
+      "step": 9600
+    },
+    {
+      "epoch": 0.22632000376807498,
+      "grad_norm": 2.220184803009033,
+      "learning_rate": 0.00015475954971503933,
+      "loss": 2.0954,
+      "step": 9610
+    },
+    {
+      "epoch": 0.22655550845461825,
+      "grad_norm": 1.9735349416732788,
+      "learning_rate": 0.00015471244877773068,
+      "loss": 2.0907,
+      "step": 9620
+    },
+    {
+      "epoch": 0.2267910131411615,
+      "grad_norm": 2.214447259902954,
+      "learning_rate": 0.00015466534784042203,
+      "loss": 2.0783,
+      "step": 9630
+    },
+    {
+      "epoch": 0.22702651782770478,
+      "grad_norm": 2.104156255722046,
+      "learning_rate": 0.00015461824690311338,
+      "loss": 2.3976,
+      "step": 9640
+    },
+    {
+      "epoch": 0.22726202251424804,
+      "grad_norm": 2.283184289932251,
+      "learning_rate": 0.00015457114596580473,
+      "loss": 2.1239,
+      "step": 9650
+    },
+    {
+      "epoch": 0.2274975272007913,
+      "grad_norm": 1.9349390268325806,
+      "learning_rate": 0.00015452404502849605,
+      "loss": 2.2178,
+      "step": 9660
+    },
+    {
+      "epoch": 0.22773303188733457,
+      "grad_norm": 2.752654552459717,
+      "learning_rate": 0.00015447694409118743,
+      "loss": 2.2175,
+      "step": 9670
+    },
+    {
+      "epoch": 0.22796853657387783,
+      "grad_norm": 1.937559723854065,
+      "learning_rate": 0.00015442984315387876,
+      "loss": 2.112,
+      "step": 9680
+    },
+    {
+      "epoch": 0.22820404126042107,
+      "grad_norm": 2.8238909244537354,
+      "learning_rate": 0.00015438274221657013,
+      "loss": 2.0023,
+      "step": 9690
+    },
+    {
+      "epoch": 0.22843954594696433,
+      "grad_norm": 2.258000373840332,
+      "learning_rate": 0.00015433564127926146,
+      "loss": 1.946,
+      "step": 9700
+    },
+    {
+      "epoch": 0.2286750506335076,
+      "grad_norm": 3.562777519226074,
+      "learning_rate": 0.0001542885403419528,
+      "loss": 2.0746,
+      "step": 9710
+    },
+    {
+      "epoch": 0.22891055532005086,
+      "grad_norm": 2.3150315284729004,
+      "learning_rate": 0.00015424143940464416,
+      "loss": 2.1203,
+      "step": 9720
+    },
+    {
+      "epoch": 0.22914606000659413,
+      "grad_norm": 3.4082276821136475,
+      "learning_rate": 0.0001541943384673355,
+      "loss": 2.3191,
+      "step": 9730
+    },
+    {
+      "epoch": 0.2293815646931374,
+      "grad_norm": 2.0063388347625732,
+      "learning_rate": 0.00015414723753002686,
+      "loss": 2.1138,
+      "step": 9740
+    },
+    {
+      "epoch": 0.22961706937968066,
+      "grad_norm": 1.8203511238098145,
+      "learning_rate": 0.0001541001365927182,
+      "loss": 2.0288,
+      "step": 9750
+    },
+    {
+      "epoch": 0.22985257406622392,
+      "grad_norm": 2.3845832347869873,
+      "learning_rate": 0.00015405303565540956,
+      "loss": 2.1712,
+      "step": 9760
+    },
+    {
+      "epoch": 0.23008807875276718,
+      "grad_norm": 1.6893136501312256,
+      "learning_rate": 0.0001540059347181009,
+      "loss": 2.215,
+      "step": 9770
+    },
+    {
+      "epoch": 0.23032358343931045,
+      "grad_norm": 2.3593993186950684,
+      "learning_rate": 0.00015395883378079226,
+      "loss": 2.081,
+      "step": 9780
+    },
+    {
+      "epoch": 0.2305590881258537,
+      "grad_norm": 2.105539321899414,
+      "learning_rate": 0.00015391173284348358,
+      "loss": 2.1697,
+      "step": 9790
+    },
+    {
+      "epoch": 0.23079459281239698,
+      "grad_norm": 1.891050100326538,
+      "learning_rate": 0.00015386463190617496,
+      "loss": 1.9764,
+      "step": 9800
+    },
+    {
+      "epoch": 0.23103009749894024,
+      "grad_norm": 4.2726263999938965,
+      "learning_rate": 0.00015381753096886628,
+      "loss": 2.3325,
+      "step": 9810
+    },
+    {
+      "epoch": 0.2312656021854835,
+      "grad_norm": 1.6610772609710693,
+      "learning_rate": 0.00015377043003155763,
+      "loss": 2.3659,
+      "step": 9820
+    },
+    {
+      "epoch": 0.23150110687202674,
+      "grad_norm": 2.202704906463623,
+      "learning_rate": 0.00015372332909424898,
+      "loss": 2.1231,
+      "step": 9830
+    },
+    {
+      "epoch": 0.23173661155857,
+      "grad_norm": 1.6619096994400024,
+      "learning_rate": 0.00015367622815694033,
+      "loss": 2.0609,
+      "step": 9840
+    },
+    {
+      "epoch": 0.23197211624511327,
+      "grad_norm": 2.1274075508117676,
+      "learning_rate": 0.00015362912721963168,
+      "loss": 2.1627,
+      "step": 9850
+    },
+    {
+      "epoch": 0.23220762093165653,
+      "grad_norm": 2.303713321685791,
+      "learning_rate": 0.00015358202628232303,
+      "loss": 2.168,
+      "step": 9860
+    },
+    {
+      "epoch": 0.2324431256181998,
+      "grad_norm": 1.953169822692871,
+      "learning_rate": 0.00015353492534501436,
+      "loss": 2.0576,
+      "step": 9870
+    },
+    {
+      "epoch": 0.23267863030474306,
+      "grad_norm": 2.4101409912109375,
+      "learning_rate": 0.00015348782440770573,
+      "loss": 2.094,
+      "step": 9880
+    },
+    {
+      "epoch": 0.23291413499128633,
+      "grad_norm": 2.0362141132354736,
+      "learning_rate": 0.00015344072347039706,
+      "loss": 2.162,
+      "step": 9890
+    },
+    {
+      "epoch": 0.2331496396778296,
+      "grad_norm": 1.7066439390182495,
+      "learning_rate": 0.00015339362253308843,
+      "loss": 2.0753,
+      "step": 9900
+    },
+    {
+      "epoch": 0.23338514436437285,
+      "grad_norm": 2.0691280364990234,
+      "learning_rate": 0.00015334652159577976,
+      "loss": 2.0622,
+      "step": 9910
+    },
+    {
+      "epoch": 0.23362064905091612,
+      "grad_norm": 2.5985186100006104,
+      "learning_rate": 0.0001532994206584711,
+      "loss": 2.1825,
+      "step": 9920
+    },
+    {
+      "epoch": 0.23385615373745938,
+      "grad_norm": 1.8621069192886353,
+      "learning_rate": 0.00015325231972116246,
+      "loss": 2.3242,
+      "step": 9930
+    },
+    {
+      "epoch": 0.23409165842400265,
+      "grad_norm": 1.9276299476623535,
+      "learning_rate": 0.0001532052187838538,
+      "loss": 1.991,
+      "step": 9940
+    },
+    {
+      "epoch": 0.2343271631105459,
+      "grad_norm": 2.965895414352417,
+      "learning_rate": 0.00015315811784654516,
+      "loss": 2.0144,
+      "step": 9950
+    },
+    {
+      "epoch": 0.23456266779708917,
+      "grad_norm": 2.115572214126587,
+      "learning_rate": 0.0001531110169092365,
+      "loss": 2.3183,
+      "step": 9960
+    },
+    {
+      "epoch": 0.2347981724836324,
+      "grad_norm": 2.177835702896118,
+      "learning_rate": 0.00015306391597192783,
+      "loss": 1.9841,
+      "step": 9970
+    },
+    {
+      "epoch": 0.23503367717017568,
+      "grad_norm": 1.9404809474945068,
+      "learning_rate": 0.0001530168150346192,
+      "loss": 2.1265,
+      "step": 9980
+    },
+    {
+      "epoch": 0.23526918185671894,
+      "grad_norm": 2.3881819248199463,
+      "learning_rate": 0.00015296971409731053,
+      "loss": 2.2287,
+      "step": 9990
+    },
+    {
+      "epoch": 0.2355046865432622,
+      "grad_norm": 2.5311830043792725,
+      "learning_rate": 0.00015292261316000188,
+      "loss": 2.1156,
+      "step": 10000
+    },
+    {
+      "epoch": 0.23574019122980547,
+      "grad_norm": 1.957363247871399,
+      "learning_rate": 0.00015287551222269323,
+      "loss": 2.0976,
+      "step": 10010
+    },
+    {
+      "epoch": 0.23597569591634873,
+      "grad_norm": 2.139197587966919,
+      "learning_rate": 0.00015282841128538458,
+      "loss": 2.1844,
+      "step": 10020
+    },
+    {
+      "epoch": 0.236211200602892,
+      "grad_norm": 2.341768741607666,
+      "learning_rate": 0.00015278131034807593,
+      "loss": 2.3208,
+      "step": 10030
+    },
+    {
+      "epoch": 0.23644670528943526,
+      "grad_norm": 2.2145462036132812,
+      "learning_rate": 0.00015273420941076728,
+      "loss": 2.225,
+      "step": 10040
+    },
+    {
+      "epoch": 0.23668220997597852,
+      "grad_norm": 2.0473461151123047,
+      "learning_rate": 0.00015268710847345863,
+      "loss": 2.3011,
+      "step": 10050
+    },
+    {
+      "epoch": 0.2369177146625218,
+      "grad_norm": 2.0674173831939697,
+      "learning_rate": 0.00015264000753614998,
+      "loss": 2.0881,
+      "step": 10060
+    },
+    {
+      "epoch": 0.23715321934906505,
+      "grad_norm": 2.2629573345184326,
+      "learning_rate": 0.00015259290659884133,
+      "loss": 2.2819,
+      "step": 10070
+    },
+    {
+      "epoch": 0.23738872403560832,
+      "grad_norm": 2.272273063659668,
+      "learning_rate": 0.00015254580566153268,
+      "loss": 2.0528,
+      "step": 10080
+    },
+    {
+      "epoch": 0.23762422872215158,
+      "grad_norm": 2.418661117553711,
+      "learning_rate": 0.00015249870472422403,
+      "loss": 2.1313,
+      "step": 10090
+    },
+    {
+      "epoch": 0.23785973340869485,
+      "grad_norm": 2.5303127765655518,
+      "learning_rate": 0.00015245160378691536,
+      "loss": 2.2907,
+      "step": 10100
+    },
+    {
+      "epoch": 0.23809523809523808,
+      "grad_norm": 2.464144229888916,
+      "learning_rate": 0.00015240450284960673,
+      "loss": 1.9191,
+      "step": 10110
+    },
+    {
+      "epoch": 0.23833074278178135,
+      "grad_norm": 1.8093903064727783,
+      "learning_rate": 0.00015235740191229806,
+      "loss": 2.0152,
+      "step": 10120
+    },
+    {
+      "epoch": 0.2385662474683246,
+      "grad_norm": 1.9383940696716309,
+      "learning_rate": 0.0001523103009749894,
+      "loss": 2.1197,
+      "step": 10130
+    },
+    {
+      "epoch": 0.23880175215486787,
+      "grad_norm": 2.671933650970459,
+      "learning_rate": 0.00015226320003768076,
+      "loss": 1.9837,
+      "step": 10140
+    },
+    {
+      "epoch": 0.23903725684141114,
+      "grad_norm": 2.083766222000122,
+      "learning_rate": 0.0001522160991003721,
+      "loss": 2.0238,
+      "step": 10150
+    },
+    {
+      "epoch": 0.2392727615279544,
+      "grad_norm": 2.2363474369049072,
+      "learning_rate": 0.00015216899816306346,
+      "loss": 2.1516,
+      "step": 10160
+    },
+    {
+      "epoch": 0.23950826621449767,
+      "grad_norm": 2.276669979095459,
+      "learning_rate": 0.0001521218972257548,
+      "loss": 1.9849,
+      "step": 10170
+    },
+    {
+      "epoch": 0.23974377090104093,
+      "grad_norm": 2.018583059310913,
+      "learning_rate": 0.00015207479628844613,
+      "loss": 2.2545,
+      "step": 10180
+    },
+    {
+      "epoch": 0.2399792755875842,
+      "grad_norm": 2.3326144218444824,
+      "learning_rate": 0.0001520276953511375,
+      "loss": 2.2914,
+      "step": 10190
+    },
+    {
+      "epoch": 0.24021478027412746,
+      "grad_norm": 2.2414534091949463,
+      "learning_rate": 0.00015198059441382883,
+      "loss": 2.0861,
+      "step": 10200
+    },
+    {
+      "epoch": 0.24045028496067072,
+      "grad_norm": 2.206983804702759,
+      "learning_rate": 0.00015193349347652018,
+      "loss": 2.1644,
+      "step": 10210
+    },
+    {
+      "epoch": 0.240685789647214,
+      "grad_norm": 1.9949047565460205,
+      "learning_rate": 0.00015188639253921153,
+      "loss": 2.2007,
+      "step": 10220
+    },
+    {
+      "epoch": 0.24092129433375725,
+      "grad_norm": 1.8927348852157593,
+      "learning_rate": 0.00015183929160190288,
+      "loss": 2.0746,
+      "step": 10230
+    },
+    {
+      "epoch": 0.24115679902030052,
+      "grad_norm": 1.8133490085601807,
+      "learning_rate": 0.00015179219066459423,
+      "loss": 1.9463,
+      "step": 10240
+    },
+    {
+      "epoch": 0.24139230370684375,
+      "grad_norm": 2.7200329303741455,
+      "learning_rate": 0.00015174508972728558,
+      "loss": 2.0079,
+      "step": 10250
+    },
+    {
+      "epoch": 0.24162780839338702,
+      "grad_norm": 2.4350547790527344,
+      "learning_rate": 0.0001516979887899769,
+      "loss": 2.2918,
+      "step": 10260
+    },
+    {
+      "epoch": 0.24186331307993028,
+      "grad_norm": 1.9246656894683838,
+      "learning_rate": 0.00015165088785266828,
+      "loss": 2.117,
+      "step": 10270
+    },
+    {
+      "epoch": 0.24209881776647355,
+      "grad_norm": 2.2206227779388428,
+      "learning_rate": 0.0001516037869153596,
+      "loss": 2.225,
+      "step": 10280
+    },
+    {
+      "epoch": 0.2423343224530168,
+      "grad_norm": 2.1294167041778564,
+      "learning_rate": 0.00015155668597805098,
+      "loss": 2.0306,
+      "step": 10290
+    },
+    {
+      "epoch": 0.24256982713956007,
+      "grad_norm": 2.4349210262298584,
+      "learning_rate": 0.0001515095850407423,
+      "loss": 2.1363,
+      "step": 10300
+    },
+    {
+      "epoch": 0.24280533182610334,
+      "grad_norm": 1.8255261182785034,
+      "learning_rate": 0.00015146248410343366,
+      "loss": 2.2975,
+      "step": 10310
+    },
+    {
+      "epoch": 0.2430408365126466,
+      "grad_norm": 2.2579727172851562,
+      "learning_rate": 0.000151415383166125,
+      "loss": 2.0691,
+      "step": 10320
+    },
+    {
+      "epoch": 0.24327634119918987,
+      "grad_norm": 2.279855966567993,
+      "learning_rate": 0.00015136828222881636,
+      "loss": 2.1189,
+      "step": 10330
+    },
+    {
+      "epoch": 0.24351184588573313,
+      "grad_norm": 2.1615757942199707,
+      "learning_rate": 0.0001513211812915077,
+      "loss": 2.2606,
+      "step": 10340
+    },
+    {
+      "epoch": 0.2437473505722764,
+      "grad_norm": 2.1494510173797607,
+      "learning_rate": 0.00015127408035419906,
+      "loss": 2.0028,
+      "step": 10350
+    },
+    {
+      "epoch": 0.24398285525881966,
+      "grad_norm": 1.856366753578186,
+      "learning_rate": 0.0001512269794168904,
+      "loss": 2.015,
+      "step": 10360
+    },
+    {
+      "epoch": 0.24421835994536292,
+      "grad_norm": 2.0949535369873047,
+      "learning_rate": 0.00015117987847958176,
+      "loss": 1.9954,
+      "step": 10370
+    },
+    {
+      "epoch": 0.2444538646319062,
+      "grad_norm": 2.0382297039031982,
+      "learning_rate": 0.0001511327775422731,
+      "loss": 2.0136,
+      "step": 10380
+    },
+    {
+      "epoch": 0.24468936931844942,
+      "grad_norm": 3.209582567214966,
+      "learning_rate": 0.00015108567660496443,
+      "loss": 2.2051,
+      "step": 10390
+    },
+    {
+      "epoch": 0.2449248740049927,
+      "grad_norm": 2.9670584201812744,
+      "learning_rate": 0.0001510385756676558,
+      "loss": 2.0755,
+      "step": 10400
+    },
+    {
+      "epoch": 0.24516037869153595,
+      "grad_norm": 2.0415966510772705,
+      "learning_rate": 0.00015099147473034713,
+      "loss": 2.0836,
+      "step": 10410
+    },
+    {
+      "epoch": 0.24539588337807922,
+      "grad_norm": 2.374818801879883,
+      "learning_rate": 0.00015094437379303848,
+      "loss": 2.1602,
+      "step": 10420
+    },
+    {
+      "epoch": 0.24563138806462248,
+      "grad_norm": 2.551715850830078,
+      "learning_rate": 0.00015089727285572983,
+      "loss": 2.2528,
+      "step": 10430
+    },
+    {
+      "epoch": 0.24586689275116574,
+      "grad_norm": 1.8446077108383179,
+      "learning_rate": 0.00015085017191842118,
+      "loss": 2.2517,
+      "step": 10440
+    },
+    {
+      "epoch": 0.246102397437709,
+      "grad_norm": 2.6373045444488525,
+      "learning_rate": 0.00015080307098111253,
+      "loss": 2.0872,
+      "step": 10450
+    },
+    {
+      "epoch": 0.24633790212425227,
+      "grad_norm": 2.356964588165283,
+      "learning_rate": 0.00015075597004380388,
+      "loss": 2.0906,
+      "step": 10460
+    },
+    {
+      "epoch": 0.24657340681079554,
+      "grad_norm": 2.3910388946533203,
+      "learning_rate": 0.0001507088691064952,
+      "loss": 2.2638,
+      "step": 10470
+    },
+    {
+      "epoch": 0.2468089114973388,
+      "grad_norm": 2.311750650405884,
+      "learning_rate": 0.00015066176816918659,
+      "loss": 1.9749,
+      "step": 10480
+    },
+    {
+      "epoch": 0.24704441618388207,
+      "grad_norm": 2.235476016998291,
+      "learning_rate": 0.0001506146672318779,
+      "loss": 1.8174,
+      "step": 10490
+    },
+    {
+      "epoch": 0.24727992087042533,
+      "grad_norm": 2.281172275543213,
+      "learning_rate": 0.00015056756629456929,
+      "loss": 1.8255,
+      "step": 10500
+    },
+    {
+      "epoch": 0.2475154255569686,
+      "grad_norm": 2.279736042022705,
+      "learning_rate": 0.0001505204653572606,
+      "loss": 2.1736,
+      "step": 10510
+    },
+    {
+      "epoch": 0.24775093024351186,
+      "grad_norm": 2.279602289199829,
+      "learning_rate": 0.00015047336441995196,
+      "loss": 2.0634,
+      "step": 10520
+    },
+    {
+      "epoch": 0.24798643493005512,
+      "grad_norm": 2.272345542907715,
+      "learning_rate": 0.0001504262634826433,
+      "loss": 2.2241,
+      "step": 10530
+    },
+    {
+      "epoch": 0.24822193961659836,
+      "grad_norm": 1.9404467344284058,
+      "learning_rate": 0.00015037916254533466,
+      "loss": 2.2235,
+      "step": 10540
+    },
+    {
+      "epoch": 0.24845744430314162,
+      "grad_norm": 2.3488924503326416,
+      "learning_rate": 0.000150332061608026,
+      "loss": 2.2408,
+      "step": 10550
+    },
+    {
+      "epoch": 0.2486929489896849,
+      "grad_norm": 1.9650462865829468,
+      "learning_rate": 0.00015028496067071736,
+      "loss": 1.9652,
+      "step": 10560
+    },
+    {
+      "epoch": 0.24892845367622815,
+      "grad_norm": 2.162095785140991,
+      "learning_rate": 0.00015023785973340868,
+      "loss": 2.2118,
+      "step": 10570
+    },
+    {
+      "epoch": 0.24916395836277141,
+      "grad_norm": 2.4109175205230713,
+      "learning_rate": 0.00015019075879610006,
+      "loss": 2.1926,
+      "step": 10580
+    },
+    {
+      "epoch": 0.24939946304931468,
+      "grad_norm": 2.6191952228546143,
+      "learning_rate": 0.00015014365785879138,
+      "loss": 2.1851,
+      "step": 10590
+    },
+    {
+      "epoch": 0.24963496773585794,
+      "grad_norm": 2.78340220451355,
+      "learning_rate": 0.00015009655692148273,
+      "loss": 2.1431,
+      "step": 10600
+    },
+    {
+      "epoch": 0.2498704724224012,
+      "grad_norm": 1.8790323734283447,
+      "learning_rate": 0.0001500494559841741,
+      "loss": 2.088,
+      "step": 10610
+    },
+    {
+      "epoch": 0.25010597710894444,
+      "grad_norm": 2.0652871131896973,
+      "learning_rate": 0.00015000235504686543,
+      "loss": 2.1255,
+      "step": 10620
+    },
+    {
+      "epoch": 0.2503414817954877,
+      "grad_norm": 1.9595967531204224,
+      "learning_rate": 0.00014995525410955678,
+      "loss": 2.026,
+      "step": 10630
+    },
+    {
+      "epoch": 0.250576986482031,
+      "grad_norm": 2.293111562728882,
+      "learning_rate": 0.00014990815317224813,
+      "loss": 2.2182,
+      "step": 10640
+    },
+    {
+      "epoch": 0.25081249116857424,
+      "grad_norm": 1.931410789489746,
+      "learning_rate": 0.00014986105223493948,
+      "loss": 2.2731,
+      "step": 10650
+    },
+    {
+      "epoch": 0.2510479958551175,
+      "grad_norm": 2.6181063652038574,
+      "learning_rate": 0.00014981395129763084,
+      "loss": 2.2646,
+      "step": 10660
+    },
+    {
+      "epoch": 0.25128350054166076,
+      "grad_norm": 1.9321035146713257,
+      "learning_rate": 0.00014976685036032219,
+      "loss": 2.1447,
+      "step": 10670
+    },
+    {
+      "epoch": 0.25151900522820403,
+      "grad_norm": 3.149609327316284,
+      "learning_rate": 0.00014971974942301354,
+      "loss": 2.1219,
+      "step": 10680
+    },
+    {
+      "epoch": 0.2517545099147473,
+      "grad_norm": 1.9660497903823853,
+      "learning_rate": 0.00014967264848570489,
+      "loss": 2.2534,
+      "step": 10690
+    },
+    {
+      "epoch": 0.25199001460129056,
+      "grad_norm": 2.55802845954895,
+      "learning_rate": 0.0001496255475483962,
+      "loss": 2.2268,
+      "step": 10700
+    },
+    {
+      "epoch": 0.2522255192878338,
+      "grad_norm": 1.8749346733093262,
+      "learning_rate": 0.0001495784466110876,
+      "loss": 2.0165,
+      "step": 10710
+    },
+    {
+      "epoch": 0.2524610239743771,
+      "grad_norm": 1.9412798881530762,
+      "learning_rate": 0.0001495313456737789,
+      "loss": 2.2144,
+      "step": 10720
+    },
+    {
+      "epoch": 0.25269652866092035,
+      "grad_norm": 2.201239585876465,
+      "learning_rate": 0.00014948424473647026,
+      "loss": 2.1653,
+      "step": 10730
+    },
+    {
+      "epoch": 0.2529320333474636,
+      "grad_norm": 1.9565176963806152,
+      "learning_rate": 0.0001494371437991616,
+      "loss": 2.1649,
+      "step": 10740
+    },
+    {
+      "epoch": 0.2531675380340069,
+      "grad_norm": 2.220141649246216,
+      "learning_rate": 0.00014939004286185296,
+      "loss": 1.8715,
+      "step": 10750
+    },
+    {
+      "epoch": 0.25340304272055014,
+      "grad_norm": 2.311554193496704,
+      "learning_rate": 0.0001493429419245443,
+      "loss": 2.2611,
+      "step": 10760
+    },
+    {
+      "epoch": 0.2536385474070934,
+      "grad_norm": 2.3438363075256348,
+      "learning_rate": 0.00014929584098723566,
+      "loss": 2.1026,
+      "step": 10770
+    },
+    {
+      "epoch": 0.25387405209363667,
+      "grad_norm": 1.9707456827163696,
+      "learning_rate": 0.00014924874004992698,
+      "loss": 2.0283,
+      "step": 10780
+    },
+    {
+      "epoch": 0.25410955678017993,
+      "grad_norm": 3.091794729232788,
+      "learning_rate": 0.00014920163911261836,
+      "loss": 2.3363,
+      "step": 10790
+    },
+    {
+      "epoch": 0.2543450614667232,
+      "grad_norm": 2.3337857723236084,
+      "learning_rate": 0.00014915453817530968,
+      "loss": 2.1223,
+      "step": 10800
+    },
+    {
+      "epoch": 0.25458056615326646,
+      "grad_norm": 1.6483640670776367,
+      "learning_rate": 0.00014910743723800103,
+      "loss": 2.2096,
+      "step": 10810
+    },
+    {
+      "epoch": 0.2548160708398097,
+      "grad_norm": 2.229172945022583,
+      "learning_rate": 0.00014906033630069238,
+      "loss": 2.1877,
+      "step": 10820
+    },
+    {
+      "epoch": 0.255051575526353,
+      "grad_norm": 1.7530523538589478,
+      "learning_rate": 0.00014901323536338374,
+      "loss": 1.9589,
+      "step": 10830
+    },
+    {
+      "epoch": 0.25528708021289626,
+      "grad_norm": 1.962538242340088,
+      "learning_rate": 0.00014896613442607509,
+      "loss": 2.2333,
+      "step": 10840
+    },
+    {
+      "epoch": 0.2555225848994395,
+      "grad_norm": 3.094045877456665,
+      "learning_rate": 0.00014891903348876644,
+      "loss": 2.1465,
+      "step": 10850
+    },
+    {
+      "epoch": 0.2557580895859828,
+      "grad_norm": 1.9345427751541138,
+      "learning_rate": 0.00014887193255145776,
+      "loss": 2.1342,
+      "step": 10860
+    },
+    {
+      "epoch": 0.25599359427252605,
+      "grad_norm": 1.932407259941101,
+      "learning_rate": 0.00014882483161414914,
+      "loss": 2.0455,
+      "step": 10870
+    },
+    {
+      "epoch": 0.2562290989590693,
+      "grad_norm": 2.0643908977508545,
+      "learning_rate": 0.00014877773067684046,
+      "loss": 2.0413,
+      "step": 10880
+    },
+    {
+      "epoch": 0.2564646036456125,
+      "grad_norm": 2.0406455993652344,
+      "learning_rate": 0.00014873062973953184,
+      "loss": 2.2408,
+      "step": 10890
+    },
+    {
+      "epoch": 0.2567001083321558,
+      "grad_norm": 2.9993858337402344,
+      "learning_rate": 0.0001486835288022232,
+      "loss": 2.1404,
+      "step": 10900
+    },
+    {
+      "epoch": 0.25693561301869905,
+      "grad_norm": 2.122169017791748,
+      "learning_rate": 0.0001486364278649145,
+      "loss": 2.3794,
+      "step": 10910
+    },
+    {
+      "epoch": 0.2571711177052423,
+      "grad_norm": 2.0623533725738525,
+      "learning_rate": 0.0001485893269276059,
+      "loss": 2.089,
+      "step": 10920
+    },
+    {
+      "epoch": 0.2574066223917856,
+      "grad_norm": 2.2135274410247803,
+      "learning_rate": 0.0001485422259902972,
+      "loss": 2.0781,
+      "step": 10930
+    },
+    {
+      "epoch": 0.25764212707832884,
+      "grad_norm": 3.7605085372924805,
+      "learning_rate": 0.00014849512505298856,
+      "loss": 2.3425,
+      "step": 10940
+    },
+    {
+      "epoch": 0.2578776317648721,
+      "grad_norm": 2.248478412628174,
+      "learning_rate": 0.0001484480241156799,
+      "loss": 2.0831,
+      "step": 10950
+    },
+    {
+      "epoch": 0.25811313645141537,
+      "grad_norm": 2.040879249572754,
+      "learning_rate": 0.00014840092317837126,
+      "loss": 2.1721,
+      "step": 10960
+    },
+    {
+      "epoch": 0.25834864113795863,
+      "grad_norm": 2.4752190113067627,
+      "learning_rate": 0.0001483538222410626,
+      "loss": 1.8606,
+      "step": 10970
+    },
+    {
+      "epoch": 0.2585841458245019,
+      "grad_norm": 2.079008102416992,
+      "learning_rate": 0.00014830672130375396,
+      "loss": 2.1368,
+      "step": 10980
+    },
+    {
+      "epoch": 0.25881965051104516,
+      "grad_norm": 2.152157783508301,
+      "learning_rate": 0.00014825962036644528,
+      "loss": 2.2405,
+      "step": 10990
+    },
+    {
+      "epoch": 0.2590551551975884,
+      "grad_norm": 1.8461806774139404,
+      "learning_rate": 0.00014821251942913666,
+      "loss": 2.2071,
+      "step": 11000
+    },
+    {
+      "epoch": 0.2592906598841317,
+      "grad_norm": 2.190112590789795,
+      "learning_rate": 0.00014816541849182799,
+      "loss": 2.1474,
+      "step": 11010
+    },
+    {
+      "epoch": 0.25952616457067496,
+      "grad_norm": 1.828783631324768,
+      "learning_rate": 0.00014811831755451934,
+      "loss": 2.0584,
+      "step": 11020
+    },
+    {
+      "epoch": 0.2597616692572182,
+      "grad_norm": 2.171999454498291,
+      "learning_rate": 0.00014807121661721069,
+      "loss": 2.2661,
+      "step": 11030
+    },
+    {
+      "epoch": 0.2599971739437615,
+      "grad_norm": 2.530381679534912,
+      "learning_rate": 0.00014802411567990204,
+      "loss": 2.3015,
+      "step": 11040
+    },
+    {
+      "epoch": 0.26023267863030475,
+      "grad_norm": 2.840162515640259,
+      "learning_rate": 0.00014797701474259339,
+      "loss": 2.0476,
+      "step": 11050
+    },
+    {
+      "epoch": 0.260468183316848,
+      "grad_norm": 2.931011438369751,
+      "learning_rate": 0.00014792991380528474,
+      "loss": 2.1905,
+      "step": 11060
+    },
+    {
+      "epoch": 0.2607036880033913,
+      "grad_norm": 2.2790048122406006,
+      "learning_rate": 0.00014788281286797606,
+      "loss": 1.9775,
+      "step": 11070
+    },
+    {
+      "epoch": 0.26093919268993454,
+      "grad_norm": 2.563983678817749,
+      "learning_rate": 0.00014783571193066744,
+      "loss": 2.0517,
+      "step": 11080
+    },
+    {
+      "epoch": 0.2611746973764778,
+      "grad_norm": 1.9197920560836792,
+      "learning_rate": 0.00014779332108708963,
+      "loss": 2.0083,
+      "step": 11090
+    },
+    {
+      "epoch": 0.26141020206302107,
+      "grad_norm": 2.238065242767334,
+      "learning_rate": 0.00014774622014978098,
+      "loss": 2.1118,
+      "step": 11100
+    },
+    {
+      "epoch": 0.26164570674956433,
+      "grad_norm": 2.00567626953125,
+      "learning_rate": 0.00014769911921247233,
+      "loss": 2.0841,
+      "step": 11110
+    },
+    {
+      "epoch": 0.2618812114361076,
+      "grad_norm": 1.9711300134658813,
+      "learning_rate": 0.00014765201827516368,
+      "loss": 2.1229,
+      "step": 11120
+    },
+    {
+      "epoch": 0.26211671612265086,
+      "grad_norm": 2.36851167678833,
+      "learning_rate": 0.00014760491733785503,
+      "loss": 1.9788,
+      "step": 11130
+    },
+    {
+      "epoch": 0.2623522208091941,
+      "grad_norm": 2.2312638759613037,
+      "learning_rate": 0.00014755781640054638,
+      "loss": 2.1221,
+      "step": 11140
+    },
+    {
+      "epoch": 0.2625877254957374,
+      "grad_norm": 1.9578697681427002,
+      "learning_rate": 0.00014751071546323773,
+      "loss": 1.9844,
+      "step": 11150
+    },
+    {
+      "epoch": 0.26282323018228065,
+      "grad_norm": 2.0937769412994385,
+      "learning_rate": 0.00014746361452592908,
+      "loss": 2.1305,
+      "step": 11160
+    },
+    {
+      "epoch": 0.26305873486882386,
+      "grad_norm": 1.9958505630493164,
+      "learning_rate": 0.00014741651358862043,
+      "loss": 2.2201,
+      "step": 11170
+    },
+    {
+      "epoch": 0.2632942395553671,
+      "grad_norm": 2.8454930782318115,
+      "learning_rate": 0.00014736941265131178,
+      "loss": 2.1756,
+      "step": 11180
+    },
+    {
+      "epoch": 0.2635297442419104,
+      "grad_norm": 1.7872885465621948,
+      "learning_rate": 0.00014732231171400313,
+      "loss": 1.788,
+      "step": 11190
+    },
+    {
+      "epoch": 0.26376524892845365,
+      "grad_norm": 2.458237886428833,
+      "learning_rate": 0.00014727521077669445,
+      "loss": 2.1815,
+      "step": 11200
+    },
+    {
+      "epoch": 0.2640007536149969,
+      "grad_norm": 2.0657920837402344,
+      "learning_rate": 0.00014722810983938583,
+      "loss": 2.0694,
+      "step": 11210
+    },
+    {
+      "epoch": 0.2642362583015402,
+      "grad_norm": 2.7617380619049072,
+      "learning_rate": 0.00014718100890207715,
+      "loss": 2.196,
+      "step": 11220
+    },
+    {
+      "epoch": 0.26447176298808345,
+      "grad_norm": 2.043877601623535,
+      "learning_rate": 0.0001471339079647685,
+      "loss": 2.2654,
+      "step": 11230
+    },
+    {
+      "epoch": 0.2647072676746267,
+      "grad_norm": 2.868586540222168,
+      "learning_rate": 0.00014708680702745985,
+      "loss": 2.1117,
+      "step": 11240
+    },
+    {
+      "epoch": 0.26494277236117,
+      "grad_norm": 2.223196268081665,
+      "learning_rate": 0.0001470397060901512,
+      "loss": 2.3975,
+      "step": 11250
+    },
+    {
+      "epoch": 0.26517827704771324,
+      "grad_norm": 2.0349698066711426,
+      "learning_rate": 0.00014699260515284255,
+      "loss": 2.047,
+      "step": 11260
+    },
+    {
+      "epoch": 0.2654137817342565,
+      "grad_norm": 1.827580451965332,
+      "learning_rate": 0.0001469455042155339,
+      "loss": 2.0944,
+      "step": 11270
+    },
+    {
+      "epoch": 0.26564928642079977,
+      "grad_norm": 2.8184313774108887,
+      "learning_rate": 0.00014689840327822523,
+      "loss": 2.0817,
+      "step": 11280
+    },
+    {
+      "epoch": 0.26588479110734303,
+      "grad_norm": 2.1076481342315674,
+      "learning_rate": 0.0001468513023409166,
+      "loss": 2.153,
+      "step": 11290
+    },
+    {
+      "epoch": 0.2661202957938863,
+      "grad_norm": 1.8298046588897705,
+      "learning_rate": 0.00014680420140360793,
+      "loss": 2.2125,
+      "step": 11300
+    },
+    {
+      "epoch": 0.26635580048042956,
+      "grad_norm": 1.8453890085220337,
+      "learning_rate": 0.00014675710046629928,
+      "loss": 2.1606,
+      "step": 11310
+    },
+    {
+      "epoch": 0.2665913051669728,
+      "grad_norm": 2.0864975452423096,
+      "learning_rate": 0.00014670999952899063,
+      "loss": 2.0076,
+      "step": 11320
+    },
+    {
+      "epoch": 0.2668268098535161,
+      "grad_norm": 2.20062255859375,
+      "learning_rate": 0.00014666289859168198,
+      "loss": 1.9654,
+      "step": 11330
+    },
+    {
+      "epoch": 0.26706231454005935,
+      "grad_norm": 1.914099097251892,
+      "learning_rate": 0.00014661579765437333,
+      "loss": 2.0639,
+      "step": 11340
+    },
+    {
+      "epoch": 0.2672978192266026,
+      "grad_norm": 1.8117254972457886,
+      "learning_rate": 0.00014656869671706468,
+      "loss": 2.0437,
+      "step": 11350
+    },
+    {
+      "epoch": 0.2675333239131459,
+      "grad_norm": 2.076619863510132,
+      "learning_rate": 0.000146521595779756,
+      "loss": 2.0552,
+      "step": 11360
+    },
+    {
+      "epoch": 0.26776882859968915,
+      "grad_norm": 1.752306580543518,
+      "learning_rate": 0.00014647449484244738,
+      "loss": 1.9671,
+      "step": 11370
+    },
+    {
+      "epoch": 0.2680043332862324,
+      "grad_norm": 2.3570029735565186,
+      "learning_rate": 0.0001464273939051387,
+      "loss": 1.8996,
+      "step": 11380
+    },
+    {
+      "epoch": 0.2682398379727757,
+      "grad_norm": 1.9335840940475464,
+      "learning_rate": 0.00014638029296783008,
+      "loss": 2.2257,
+      "step": 11390
+    },
+    {
+      "epoch": 0.26847534265931894,
+      "grad_norm": 2.8043274879455566,
+      "learning_rate": 0.0001463331920305214,
+      "loss": 2.3472,
+      "step": 11400
+    },
+    {
+      "epoch": 0.2687108473458622,
+      "grad_norm": 1.8869709968566895,
+      "learning_rate": 0.00014628609109321275,
+      "loss": 2.3359,
+      "step": 11410
+    },
+    {
+      "epoch": 0.26894635203240547,
+      "grad_norm": 3.413790225982666,
+      "learning_rate": 0.0001462389901559041,
+      "loss": 2.2504,
+      "step": 11420
+    },
+    {
+      "epoch": 0.26918185671894873,
+      "grad_norm": 2.251716136932373,
+      "learning_rate": 0.00014619188921859545,
+      "loss": 2.1951,
+      "step": 11430
+    },
+    {
+      "epoch": 0.269417361405492,
+      "grad_norm": 2.0624914169311523,
+      "learning_rate": 0.0001461447882812868,
+      "loss": 2.0907,
+      "step": 11440
+    },
+    {
+      "epoch": 0.2696528660920352,
+      "grad_norm": 2.2086572647094727,
+      "learning_rate": 0.00014609768734397816,
+      "loss": 2.0458,
+      "step": 11450
+    },
+    {
+      "epoch": 0.26988837077857847,
+      "grad_norm": 2.272899627685547,
+      "learning_rate": 0.0001460505864066695,
+      "loss": 2.0175,
+      "step": 11460
+    },
+    {
+      "epoch": 0.27012387546512173,
+      "grad_norm": 1.962472677230835,
+      "learning_rate": 0.00014600348546936086,
+      "loss": 2.2018,
+      "step": 11470
+    },
+    {
+      "epoch": 0.270359380151665,
+      "grad_norm": 1.8349707126617432,
+      "learning_rate": 0.0001459563845320522,
+      "loss": 2.1957,
+      "step": 11480
+    },
+    {
+      "epoch": 0.27059488483820826,
+      "grad_norm": 2.317610740661621,
+      "learning_rate": 0.00014590928359474353,
+      "loss": 2.1421,
+      "step": 11490
+    },
+    {
+      "epoch": 0.2708303895247515,
+      "grad_norm": 3.183609962463379,
+      "learning_rate": 0.0001458621826574349,
+      "loss": 2.0522,
+      "step": 11500
+    },
+    {
+      "epoch": 0.2710658942112948,
+      "grad_norm": 2.0786678791046143,
+      "learning_rate": 0.00014581508172012623,
+      "loss": 2.0592,
+      "step": 11510
+    },
+    {
+      "epoch": 0.27130139889783805,
+      "grad_norm": 1.7079672813415527,
+      "learning_rate": 0.00014576798078281758,
+      "loss": 1.9237,
+      "step": 11520
+    },
+    {
+      "epoch": 0.2715369035843813,
+      "grad_norm": 1.9276167154312134,
+      "learning_rate": 0.00014572087984550893,
+      "loss": 2.4222,
+      "step": 11530
+    },
+    {
+      "epoch": 0.2717724082709246,
+      "grad_norm": 2.461214303970337,
+      "learning_rate": 0.00014567377890820028,
+      "loss": 2.2114,
+      "step": 11540
+    },
+    {
+      "epoch": 0.27200791295746785,
+      "grad_norm": 1.96316659450531,
+      "learning_rate": 0.00014562667797089163,
+      "loss": 2.2535,
+      "step": 11550
+    },
+    {
+      "epoch": 0.2722434176440111,
+      "grad_norm": 2.1543939113616943,
+      "learning_rate": 0.00014557957703358298,
+      "loss": 2.2106,
+      "step": 11560
+    },
+    {
+      "epoch": 0.2724789223305544,
+      "grad_norm": 1.8701822757720947,
+      "learning_rate": 0.0001455324760962743,
+      "loss": 1.9426,
+      "step": 11570
+    },
+    {
+      "epoch": 0.27271442701709764,
+      "grad_norm": 2.2132885456085205,
+      "learning_rate": 0.00014548537515896568,
+      "loss": 2.1604,
+      "step": 11580
+    },
+    {
+      "epoch": 0.2729499317036409,
+      "grad_norm": 2.094003200531006,
+      "learning_rate": 0.000145438274221657,
+      "loss": 2.1061,
+      "step": 11590
+    },
+    {
+      "epoch": 0.27318543639018417,
+      "grad_norm": 2.395878314971924,
+      "learning_rate": 0.00014539117328434838,
+      "loss": 2.0564,
+      "step": 11600
+    },
+    {
+      "epoch": 0.27342094107672743,
+      "grad_norm": 2.653024911880493,
+      "learning_rate": 0.0001453440723470397,
+      "loss": 2.3041,
+      "step": 11610
+    },
+    {
+      "epoch": 0.2736564457632707,
+      "grad_norm": 2.932143449783325,
+      "learning_rate": 0.00014529697140973106,
+      "loss": 1.9723,
+      "step": 11620
+    },
+    {
+      "epoch": 0.27389195044981396,
+      "grad_norm": 2.500082492828369,
+      "learning_rate": 0.0001452498704724224,
+      "loss": 1.9417,
+      "step": 11630
+    },
+    {
+      "epoch": 0.2741274551363572,
+      "grad_norm": 3.0111424922943115,
+      "learning_rate": 0.00014520276953511376,
+      "loss": 2.1019,
+      "step": 11640
+    },
+    {
+      "epoch": 0.2743629598229005,
+      "grad_norm": 2.4497432708740234,
+      "learning_rate": 0.0001451556685978051,
+      "loss": 2.0508,
+      "step": 11650
+    },
+    {
+      "epoch": 0.27459846450944375,
+      "grad_norm": 2.2490170001983643,
+      "learning_rate": 0.00014510856766049646,
+      "loss": 2.1141,
+      "step": 11660
+    },
+    {
+      "epoch": 0.274833969195987,
+      "grad_norm": 2.196950674057007,
+      "learning_rate": 0.00014506146672318778,
+      "loss": 2.0565,
+      "step": 11670
+    },
+    {
+      "epoch": 0.2750694738825303,
+      "grad_norm": 3.2146339416503906,
+      "learning_rate": 0.00014501436578587916,
+      "loss": 2.0995,
+      "step": 11680
+    },
+    {
+      "epoch": 0.27530497856907354,
+      "grad_norm": 2.50426983833313,
+      "learning_rate": 0.00014496726484857048,
+      "loss": 2.166,
+      "step": 11690
+    },
+    {
+      "epoch": 0.2755404832556168,
+      "grad_norm": 2.0356388092041016,
+      "learning_rate": 0.00014492016391126183,
+      "loss": 2.1584,
+      "step": 11700
+    },
+    {
+      "epoch": 0.2757759879421601,
+      "grad_norm": 2.600349187850952,
+      "learning_rate": 0.0001448730629739532,
+      "loss": 2.1718,
+      "step": 11710
+    },
+    {
+      "epoch": 0.27601149262870334,
+      "grad_norm": 3.562185764312744,
+      "learning_rate": 0.00014482596203664453,
+      "loss": 2.2295,
+      "step": 11720
+    },
+    {
+      "epoch": 0.2762469973152466,
+      "grad_norm": 1.928037405014038,
+      "learning_rate": 0.00014477886109933588,
+      "loss": 1.9995,
+      "step": 11730
+    },
+    {
+      "epoch": 0.2764825020017898,
+      "grad_norm": 1.4152915477752686,
+      "learning_rate": 0.00014473176016202723,
+      "loss": 2.1623,
+      "step": 11740
+    },
+    {
+      "epoch": 0.2767180066883331,
+      "grad_norm": 1.9397757053375244,
+      "learning_rate": 0.00014468465922471858,
+      "loss": 2.0217,
+      "step": 11750
+    },
+    {
+      "epoch": 0.27695351137487634,
+      "grad_norm": 1.7832798957824707,
+      "learning_rate": 0.00014463755828740993,
+      "loss": 2.0114,
+      "step": 11760
+    },
+    {
+      "epoch": 0.2771890160614196,
+      "grad_norm": 2.282412528991699,
+      "learning_rate": 0.00014459045735010128,
+      "loss": 2.1579,
+      "step": 11770
+    },
+    {
+      "epoch": 0.27742452074796287,
+      "grad_norm": 2.4182028770446777,
+      "learning_rate": 0.00014454335641279263,
+      "loss": 2.0862,
+      "step": 11780
+    },
+    {
+      "epoch": 0.27766002543450613,
+      "grad_norm": 2.954577922821045,
+      "learning_rate": 0.00014449625547548398,
+      "loss": 2.184,
+      "step": 11790
+    },
+    {
+      "epoch": 0.2778955301210494,
+      "grad_norm": 1.8422027826309204,
+      "learning_rate": 0.0001444491545381753,
+      "loss": 2.1267,
+      "step": 11800
+    },
+    {
+      "epoch": 0.27813103480759266,
+      "grad_norm": 1.9264764785766602,
+      "learning_rate": 0.00014440205360086668,
+      "loss": 2.2146,
+      "step": 11810
+    },
+    {
+      "epoch": 0.2783665394941359,
+      "grad_norm": 1.719955563545227,
+      "learning_rate": 0.000144354952663558,
+      "loss": 1.9979,
+      "step": 11820
+    },
+    {
+      "epoch": 0.2786020441806792,
+      "grad_norm": 2.0132710933685303,
+      "learning_rate": 0.00014430785172624936,
+      "loss": 2.1951,
+      "step": 11830
+    },
+    {
+      "epoch": 0.27883754886722245,
+      "grad_norm": 2.1974024772644043,
+      "learning_rate": 0.0001442607507889407,
+      "loss": 2.3713,
+      "step": 11840
+    },
+    {
+      "epoch": 0.2790730535537657,
+      "grad_norm": 2.18841814994812,
+      "learning_rate": 0.00014421364985163206,
+      "loss": 2.1615,
+      "step": 11850
+    },
+    {
+      "epoch": 0.279308558240309,
+      "grad_norm": 2.1697840690612793,
+      "learning_rate": 0.0001441665489143234,
+      "loss": 2.1027,
+      "step": 11860
+    },
+    {
+      "epoch": 0.27954406292685224,
+      "grad_norm": 2.2310283184051514,
+      "learning_rate": 0.00014411944797701476,
+      "loss": 2.3225,
+      "step": 11870
+    },
+    {
+      "epoch": 0.2797795676133955,
+      "grad_norm": 2.3003742694854736,
+      "learning_rate": 0.00014407234703970608,
+      "loss": 2.1095,
+      "step": 11880
+    },
+    {
+      "epoch": 0.28001507229993877,
+      "grad_norm": 1.808599591255188,
+      "learning_rate": 0.00014402524610239746,
+      "loss": 2.0941,
+      "step": 11890
+    },
+    {
+      "epoch": 0.28025057698648204,
+      "grad_norm": 2.0506701469421387,
+      "learning_rate": 0.00014397814516508878,
+      "loss": 2.3607,
+      "step": 11900
+    },
+    {
+      "epoch": 0.2804860816730253,
+      "grad_norm": 1.990747332572937,
+      "learning_rate": 0.00014393104422778013,
+      "loss": 2.0992,
+      "step": 11910
+    },
+    {
+      "epoch": 0.28072158635956856,
+      "grad_norm": 2.9683446884155273,
+      "learning_rate": 0.00014388394329047148,
+      "loss": 1.9899,
+      "step": 11920
+    },
+    {
+      "epoch": 0.28095709104611183,
+      "grad_norm": 2.2459323406219482,
+      "learning_rate": 0.00014383684235316283,
+      "loss": 1.9907,
+      "step": 11930
+    },
+    {
+      "epoch": 0.2811925957326551,
+      "grad_norm": 2.313927173614502,
+      "learning_rate": 0.00014378974141585418,
+      "loss": 2.1981,
+      "step": 11940
+    },
+    {
+      "epoch": 0.28142810041919836,
+      "grad_norm": 2.565319299697876,
+      "learning_rate": 0.00014374264047854553,
+      "loss": 2.2282,
+      "step": 11950
+    },
+    {
+      "epoch": 0.2816636051057416,
+      "grad_norm": 2.329613208770752,
+      "learning_rate": 0.00014369553954123686,
+      "loss": 2.0866,
+      "step": 11960
+    },
+    {
+      "epoch": 0.2818991097922849,
+      "grad_norm": 1.7780147790908813,
+      "learning_rate": 0.00014364843860392823,
+      "loss": 2.1109,
+      "step": 11970
+    },
+    {
+      "epoch": 0.28213461447882815,
+      "grad_norm": 2.3526809215545654,
+      "learning_rate": 0.00014360133766661956,
+      "loss": 2.0509,
+      "step": 11980
+    },
+    {
+      "epoch": 0.2823701191653714,
+      "grad_norm": 2.9026174545288086,
+      "learning_rate": 0.00014355423672931093,
+      "loss": 2.1536,
+      "step": 11990
+    },
+    {
+      "epoch": 0.2826056238519147,
+      "grad_norm": 2.3151416778564453,
+      "learning_rate": 0.00014350713579200228,
+      "loss": 2.2772,
+      "step": 12000
+    },
+    {
+      "epoch": 0.28284112853845794,
+      "grad_norm": 2.0522875785827637,
+      "learning_rate": 0.0001434600348546936,
+      "loss": 1.9419,
+      "step": 12010
+    },
+    {
+      "epoch": 0.28307663322500115,
+      "grad_norm": 1.7886687517166138,
+      "learning_rate": 0.00014341293391738498,
+      "loss": 2.1908,
+      "step": 12020
+    },
+    {
+      "epoch": 0.2833121379115444,
+      "grad_norm": 2.5073275566101074,
+      "learning_rate": 0.0001433658329800763,
+      "loss": 2.2582,
+      "step": 12030
+    },
+    {
+      "epoch": 0.2835476425980877,
+      "grad_norm": 1.9218273162841797,
+      "learning_rate": 0.00014331873204276766,
+      "loss": 2.1327,
+      "step": 12040
+    },
+    {
+      "epoch": 0.28378314728463094,
+      "grad_norm": 1.9071509838104248,
+      "learning_rate": 0.000143271631105459,
+      "loss": 2.1697,
+      "step": 12050
+    },
+    {
+      "epoch": 0.2840186519711742,
+      "grad_norm": 1.9115699529647827,
+      "learning_rate": 0.00014322453016815036,
+      "loss": 2.1689,
+      "step": 12060
+    },
+    {
+      "epoch": 0.28425415665771747,
+      "grad_norm": 3.1587018966674805,
+      "learning_rate": 0.0001431774292308417,
+      "loss": 2.2849,
+      "step": 12070
+    },
+    {
+      "epoch": 0.28448966134426074,
+      "grad_norm": 2.1622958183288574,
+      "learning_rate": 0.00014313032829353306,
+      "loss": 1.9378,
+      "step": 12080
+    },
+    {
+      "epoch": 0.284725166030804,
+      "grad_norm": 2.525998830795288,
+      "learning_rate": 0.00014308322735622438,
+      "loss": 2.277,
+      "step": 12090
+    },
+    {
+      "epoch": 0.28496067071734726,
+      "grad_norm": 1.5249601602554321,
+      "learning_rate": 0.00014303612641891576,
+      "loss": 1.9813,
+      "step": 12100
+    },
+    {
+      "epoch": 0.28519617540389053,
+      "grad_norm": 2.085453748703003,
+      "learning_rate": 0.00014298902548160708,
+      "loss": 2.3223,
+      "step": 12110
+    },
+    {
+      "epoch": 0.2854316800904338,
+      "grad_norm": 1.8140358924865723,
+      "learning_rate": 0.00014294192454429843,
+      "loss": 2.1393,
+      "step": 12120
+    },
+    {
+      "epoch": 0.28566718477697706,
+      "grad_norm": 2.5443224906921387,
+      "learning_rate": 0.00014289482360698978,
+      "loss": 2.2873,
+      "step": 12130
+    },
+    {
+      "epoch": 0.2859026894635203,
+      "grad_norm": 2.006328821182251,
+      "learning_rate": 0.00014284772266968113,
+      "loss": 2.1113,
+      "step": 12140
+    },
+    {
+      "epoch": 0.2861381941500636,
+      "grad_norm": 1.8038783073425293,
+      "learning_rate": 0.00014280062173237248,
+      "loss": 2.3057,
+      "step": 12150
+    },
+    {
+      "epoch": 0.28637369883660685,
+      "grad_norm": 2.357002019882202,
+      "learning_rate": 0.00014275352079506383,
+      "loss": 2.0994,
+      "step": 12160
+    },
+    {
+      "epoch": 0.2866092035231501,
+      "grad_norm": 2.0927555561065674,
+      "learning_rate": 0.00014270641985775516,
+      "loss": 1.9388,
+      "step": 12170
+    },
+    {
+      "epoch": 0.2868447082096934,
+      "grad_norm": 3.126340866088867,
+      "learning_rate": 0.00014265931892044653,
+      "loss": 2.2587,
+      "step": 12180
+    },
+    {
+      "epoch": 0.28708021289623664,
+      "grad_norm": 2.3898532390594482,
+      "learning_rate": 0.00014261221798313786,
+      "loss": 2.0545,
+      "step": 12190
+    },
+    {
+      "epoch": 0.2873157175827799,
+      "grad_norm": 2.6160011291503906,
+      "learning_rate": 0.00014256511704582923,
+      "loss": 1.9734,
+      "step": 12200
+    },
+    {
+      "epoch": 0.28755122226932317,
+      "grad_norm": 2.259218692779541,
+      "learning_rate": 0.00014251801610852056,
+      "loss": 2.0344,
+      "step": 12210
+    },
+    {
+      "epoch": 0.28778672695586643,
+      "grad_norm": 2.0822460651397705,
+      "learning_rate": 0.0001424709151712119,
+      "loss": 2.2014,
+      "step": 12220
+    },
+    {
+      "epoch": 0.2880222316424097,
+      "grad_norm": 2.0669522285461426,
+      "learning_rate": 0.00014242381423390326,
+      "loss": 2.0617,
+      "step": 12230
+    },
+    {
+      "epoch": 0.28825773632895296,
+      "grad_norm": 1.9552747011184692,
+      "learning_rate": 0.0001423767132965946,
+      "loss": 2.1155,
+      "step": 12240
+    },
+    {
+      "epoch": 0.2884932410154962,
+      "grad_norm": 2.3942112922668457,
+      "learning_rate": 0.00014232961235928596,
+      "loss": 2.0574,
+      "step": 12250
+    },
+    {
+      "epoch": 0.2887287457020395,
+      "grad_norm": 2.205756187438965,
+      "learning_rate": 0.0001422825114219773,
+      "loss": 2.1403,
+      "step": 12260
+    },
+    {
+      "epoch": 0.28896425038858276,
+      "grad_norm": 2.649390697479248,
+      "learning_rate": 0.00014223541048466863,
+      "loss": 2.1289,
+      "step": 12270
+    },
+    {
+      "epoch": 0.289199755075126,
+      "grad_norm": 2.3708388805389404,
+      "learning_rate": 0.00014218830954736,
+      "loss": 2.06,
+      "step": 12280
+    },
+    {
+      "epoch": 0.2894352597616693,
+      "grad_norm": 2.051314115524292,
+      "learning_rate": 0.00014214120861005136,
+      "loss": 2.2359,
+      "step": 12290
+    },
+    {
+      "epoch": 0.2896707644482125,
+      "grad_norm": 1.698848009109497,
+      "learning_rate": 0.00014209410767274268,
+      "loss": 2.0647,
+      "step": 12300
+    },
+    {
+      "epoch": 0.28990626913475576,
+      "grad_norm": 1.9865273237228394,
+      "learning_rate": 0.00014204700673543406,
+      "loss": 1.9384,
+      "step": 12310
+    },
+    {
+      "epoch": 0.290141773821299,
+      "grad_norm": 2.2270126342773438,
+      "learning_rate": 0.00014199990579812538,
+      "loss": 2.1169,
+      "step": 12320
+    },
+    {
+      "epoch": 0.2903772785078423,
+      "grad_norm": 1.6230186223983765,
+      "learning_rate": 0.00014195280486081673,
+      "loss": 1.9432,
+      "step": 12330
+    },
+    {
+      "epoch": 0.29061278319438555,
+      "grad_norm": 1.7483242750167847,
+      "learning_rate": 0.00014190570392350808,
+      "loss": 2.1609,
+      "step": 12340
+    },
+    {
+      "epoch": 0.2908482878809288,
+      "grad_norm": 1.7598001956939697,
+      "learning_rate": 0.00014185860298619943,
+      "loss": 2.396,
+      "step": 12350
+    },
+    {
+      "epoch": 0.2910837925674721,
+      "grad_norm": 2.14711856842041,
+      "learning_rate": 0.00014181150204889078,
+      "loss": 2.092,
+      "step": 12360
+    },
+    {
+      "epoch": 0.29131929725401534,
+      "grad_norm": 1.9929486513137817,
+      "learning_rate": 0.00014176440111158213,
+      "loss": 2.0726,
+      "step": 12370
+    },
+    {
+      "epoch": 0.2915548019405586,
+      "grad_norm": 1.9438811540603638,
+      "learning_rate": 0.00014171730017427348,
+      "loss": 1.9861,
+      "step": 12380
+    },
+    {
+      "epoch": 0.29179030662710187,
+      "grad_norm": 1.812244176864624,
+      "learning_rate": 0.00014167019923696483,
+      "loss": 2.2,
+      "step": 12390
+    },
+    {
+      "epoch": 0.29202581131364513,
+      "grad_norm": 1.8583556413650513,
+      "learning_rate": 0.00014162309829965616,
+      "loss": 2.0777,
+      "step": 12400
+    },
+    {
+      "epoch": 0.2922613160001884,
+      "grad_norm": 2.8509726524353027,
+      "learning_rate": 0.00014157599736234753,
+      "loss": 2.112,
+      "step": 12410
+    },
+    {
+      "epoch": 0.29249682068673166,
+      "grad_norm": 2.370173931121826,
+      "learning_rate": 0.00014152889642503886,
+      "loss": 2.1104,
+      "step": 12420
+    },
+    {
+      "epoch": 0.2927323253732749,
+      "grad_norm": 2.105886459350586,
+      "learning_rate": 0.0001414817954877302,
+      "loss": 2.1593,
+      "step": 12430
+    },
+    {
+      "epoch": 0.2929678300598182,
+      "grad_norm": 2.6540844440460205,
+      "learning_rate": 0.00014143469455042156,
+      "loss": 2.034,
+      "step": 12440
+    },
+    {
+      "epoch": 0.29320333474636145,
+      "grad_norm": 1.7813196182250977,
+      "learning_rate": 0.0001413875936131129,
+      "loss": 1.899,
+      "step": 12450
+    },
+    {
+      "epoch": 0.2934388394329047,
+      "grad_norm": 2.747343063354492,
+      "learning_rate": 0.00014134049267580426,
+      "loss": 2.1251,
+      "step": 12460
+    },
+    {
+      "epoch": 0.293674344119448,
+      "grad_norm": 2.120857000350952,
+      "learning_rate": 0.0001412933917384956,
+      "loss": 1.9557,
+      "step": 12470
+    },
+    {
+      "epoch": 0.29390984880599125,
+      "grad_norm": 2.8175089359283447,
+      "learning_rate": 0.00014124629080118693,
+      "loss": 2.0522,
+      "step": 12480
+    },
+    {
+      "epoch": 0.2941453534925345,
+      "grad_norm": 3.126023292541504,
+      "learning_rate": 0.0001411991898638783,
+      "loss": 2.06,
+      "step": 12490
+    },
+    {
+      "epoch": 0.2943808581790778,
+      "grad_norm": 2.139693260192871,
+      "learning_rate": 0.00014115208892656963,
+      "loss": 2.2009,
+      "step": 12500
+    },
+    {
+      "epoch": 0.29461636286562104,
+      "grad_norm": 2.007183790206909,
+      "learning_rate": 0.00014110498798926098,
+      "loss": 2.1605,
+      "step": 12510
+    },
+    {
+      "epoch": 0.2948518675521643,
+      "grad_norm": 1.9379314184188843,
+      "learning_rate": 0.00014105788705195233,
+      "loss": 2.1381,
+      "step": 12520
+    },
+    {
+      "epoch": 0.29508737223870757,
+      "grad_norm": 2.096290349960327,
+      "learning_rate": 0.00014101078611464368,
+      "loss": 2.2027,
+      "step": 12530
+    },
+    {
+      "epoch": 0.29532287692525083,
+      "grad_norm": 2.452683925628662,
+      "learning_rate": 0.00014096368517733503,
+      "loss": 2.1766,
+      "step": 12540
+    },
+    {
+      "epoch": 0.2955583816117941,
+      "grad_norm": 2.5896694660186768,
+      "learning_rate": 0.00014091658424002638,
+      "loss": 2.1109,
+      "step": 12550
+    },
+    {
+      "epoch": 0.29579388629833736,
+      "grad_norm": 2.125363349914551,
+      "learning_rate": 0.00014086948330271773,
+      "loss": 2.3477,
+      "step": 12560
+    },
+    {
+      "epoch": 0.2960293909848806,
+      "grad_norm": 1.9147471189498901,
+      "learning_rate": 0.00014082238236540908,
+      "loss": 2.4074,
+      "step": 12570
+    },
+    {
+      "epoch": 0.2962648956714239,
+      "grad_norm": 2.466128349304199,
+      "learning_rate": 0.00014077528142810043,
+      "loss": 2.161,
+      "step": 12580
+    },
+    {
+      "epoch": 0.2965004003579671,
+      "grad_norm": 2.0209436416625977,
+      "learning_rate": 0.00014072818049079179,
+      "loss": 2.0503,
+      "step": 12590
+    },
+    {
+      "epoch": 0.29673590504451036,
+      "grad_norm": 2.1198041439056396,
+      "learning_rate": 0.00014068107955348314,
+      "loss": 2.166,
+      "step": 12600
+    },
+    {
+      "epoch": 0.2969714097310536,
+      "grad_norm": 1.7354202270507812,
+      "learning_rate": 0.00014063397861617446,
+      "loss": 2.0945,
+      "step": 12610
+    },
+    {
+      "epoch": 0.2972069144175969,
+      "grad_norm": 1.8669902086257935,
+      "learning_rate": 0.00014058687767886584,
+      "loss": 1.9682,
+      "step": 12620
+    },
+    {
+      "epoch": 0.29744241910414015,
+      "grad_norm": 2.1740376949310303,
+      "learning_rate": 0.00014053977674155716,
+      "loss": 2.1025,
+      "step": 12630
+    },
+    {
+      "epoch": 0.2976779237906834,
+      "grad_norm": 2.275982141494751,
+      "learning_rate": 0.0001404926758042485,
+      "loss": 2.2597,
+      "step": 12640
+    },
+    {
+      "epoch": 0.2979134284772267,
+      "grad_norm": 2.172459125518799,
+      "learning_rate": 0.00014044557486693986,
+      "loss": 2.1002,
+      "step": 12650
+    },
+    {
+      "epoch": 0.29814893316376995,
+      "grad_norm": 2.4488515853881836,
+      "learning_rate": 0.0001403984739296312,
+      "loss": 1.9405,
+      "step": 12660
+    },
+    {
+      "epoch": 0.2983844378503132,
+      "grad_norm": 2.1682851314544678,
+      "learning_rate": 0.00014035137299232256,
+      "loss": 2.1472,
+      "step": 12670
+    },
+    {
+      "epoch": 0.2986199425368565,
+      "grad_norm": 1.9250439405441284,
+      "learning_rate": 0.0001403042720550139,
+      "loss": 2.1314,
+      "step": 12680
+    },
+    {
+      "epoch": 0.29885544722339974,
+      "grad_norm": 3.761707305908203,
+      "learning_rate": 0.00014025717111770523,
+      "loss": 2.1523,
+      "step": 12690
+    },
+    {
+      "epoch": 0.299090951909943,
+      "grad_norm": 2.87754487991333,
+      "learning_rate": 0.0001402100701803966,
+      "loss": 1.9867,
+      "step": 12700
+    },
+    {
+      "epoch": 0.29932645659648627,
+      "grad_norm": 2.3228402137756348,
+      "learning_rate": 0.00014016296924308793,
+      "loss": 2.015,
+      "step": 12710
+    },
+    {
+      "epoch": 0.29956196128302953,
+      "grad_norm": 1.869566798210144,
+      "learning_rate": 0.00014011586830577928,
+      "loss": 2.2132,
+      "step": 12720
+    },
+    {
+      "epoch": 0.2997974659695728,
+      "grad_norm": 2.2412784099578857,
+      "learning_rate": 0.00014006876736847063,
+      "loss": 2.3368,
+      "step": 12730
+    },
+    {
+      "epoch": 0.30003297065611606,
+      "grad_norm": 1.911073923110962,
+      "learning_rate": 0.00014002166643116198,
+      "loss": 2.0563,
+      "step": 12740
+    },
+    {
+      "epoch": 0.3002684753426593,
+      "grad_norm": 2.371375799179077,
+      "learning_rate": 0.00013997456549385333,
+      "loss": 2.1771,
+      "step": 12750
+    },
+    {
+      "epoch": 0.3005039800292026,
+      "grad_norm": 2.312708854675293,
+      "learning_rate": 0.00013992746455654469,
+      "loss": 2.158,
+      "step": 12760
+    },
+    {
+      "epoch": 0.30073948471574585,
+      "grad_norm": 2.721620559692383,
+      "learning_rate": 0.000139880363619236,
+      "loss": 2.2242,
+      "step": 12770
+    },
+    {
+      "epoch": 0.3009749894022891,
+      "grad_norm": 2.6075215339660645,
+      "learning_rate": 0.00013983326268192739,
+      "loss": 2.029,
+      "step": 12780
+    },
+    {
+      "epoch": 0.3012104940888324,
+      "grad_norm": 2.001995086669922,
+      "learning_rate": 0.0001397861617446187,
+      "loss": 2.1364,
+      "step": 12790
+    },
+    {
+      "epoch": 0.30144599877537565,
+      "grad_norm": 1.792315125465393,
+      "learning_rate": 0.00013973906080731009,
+      "loss": 2.0791,
+      "step": 12800
+    },
+    {
+      "epoch": 0.3016815034619189,
+      "grad_norm": 2.6156537532806396,
+      "learning_rate": 0.0001396919598700014,
+      "loss": 2.131,
+      "step": 12810
+    },
+    {
+      "epoch": 0.3019170081484622,
+      "grad_norm": 2.277663230895996,
+      "learning_rate": 0.00013964485893269276,
+      "loss": 2.03,
+      "step": 12820
+    },
+    {
+      "epoch": 0.30215251283500544,
+      "grad_norm": 2.09621000289917,
+      "learning_rate": 0.0001395977579953841,
+      "loss": 2.2744,
+      "step": 12830
+    },
+    {
+      "epoch": 0.3023880175215487,
+      "grad_norm": 2.4095897674560547,
+      "learning_rate": 0.00013955065705807546,
+      "loss": 1.8946,
+      "step": 12840
+    },
+    {
+      "epoch": 0.30262352220809197,
+      "grad_norm": 2.2622735500335693,
+      "learning_rate": 0.0001395035561207668,
+      "loss": 2.1998,
+      "step": 12850
+    },
+    {
+      "epoch": 0.30285902689463523,
+      "grad_norm": 2.8324317932128906,
+      "learning_rate": 0.00013945645518345816,
+      "loss": 2.4248,
+      "step": 12860
+    },
+    {
+      "epoch": 0.30309453158117844,
+      "grad_norm": 2.2162587642669678,
+      "learning_rate": 0.0001394093542461495,
+      "loss": 2.0623,
+      "step": 12870
+    },
+    {
+      "epoch": 0.3033300362677217,
+      "grad_norm": 2.325887441635132,
+      "learning_rate": 0.00013936225330884086,
+      "loss": 2.2047,
+      "step": 12880
+    },
+    {
+      "epoch": 0.30356554095426497,
+      "grad_norm": 2.152468681335449,
+      "learning_rate": 0.0001393151523715322,
+      "loss": 2.1984,
+      "step": 12890
+    },
+    {
+      "epoch": 0.30380104564080823,
+      "grad_norm": 2.0017037391662598,
+      "learning_rate": 0.00013926805143422353,
+      "loss": 2.1574,
+      "step": 12900
+    },
+    {
+      "epoch": 0.3040365503273515,
+      "grad_norm": 2.003835439682007,
+      "learning_rate": 0.0001392209504969149,
+      "loss": 2.2094,
+      "step": 12910
+    },
+    {
+      "epoch": 0.30427205501389476,
+      "grad_norm": 2.4444828033447266,
+      "learning_rate": 0.00013917384955960623,
+      "loss": 2.0709,
+      "step": 12920
+    },
+    {
+      "epoch": 0.304507559700438,
+      "grad_norm": 2.4966955184936523,
+      "learning_rate": 0.00013912674862229758,
+      "loss": 2.0336,
+      "step": 12930
+    },
+    {
+      "epoch": 0.3047430643869813,
+      "grad_norm": 2.560464859008789,
+      "learning_rate": 0.00013907964768498894,
+      "loss": 2.0573,
+      "step": 12940
+    },
+    {
+      "epoch": 0.30497856907352455,
+      "grad_norm": 2.135472059249878,
+      "learning_rate": 0.00013903254674768029,
+      "loss": 2.0721,
+      "step": 12950
+    },
+    {
+      "epoch": 0.3052140737600678,
+      "grad_norm": 1.9246833324432373,
+      "learning_rate": 0.00013898544581037164,
+      "loss": 2.1015,
+      "step": 12960
+    },
+    {
+      "epoch": 0.3054495784466111,
+      "grad_norm": 2.2237212657928467,
+      "learning_rate": 0.00013893834487306299,
+      "loss": 1.9604,
+      "step": 12970
+    },
+    {
+      "epoch": 0.30568508313315434,
+      "grad_norm": 2.493617057800293,
+      "learning_rate": 0.00013889124393575434,
+      "loss": 2.278,
+      "step": 12980
+    },
+    {
+      "epoch": 0.3059205878196976,
+      "grad_norm": 3.1759326457977295,
+      "learning_rate": 0.00013884414299844569,
+      "loss": 2.1974,
+      "step": 12990
+    },
+    {
+      "epoch": 0.3061560925062409,
+      "grad_norm": 2.2904551029205322,
+      "learning_rate": 0.000138797042061137,
+      "loss": 2.0009,
+      "step": 13000
+    },
+    {
+      "epoch": 0.30639159719278414,
+      "grad_norm": 2.1130900382995605,
+      "learning_rate": 0.0001387499411238284,
+      "loss": 2.1561,
+      "step": 13010
+    },
+    {
+      "epoch": 0.3066271018793274,
+      "grad_norm": 2.156583786010742,
+      "learning_rate": 0.0001387028401865197,
+      "loss": 2.1903,
+      "step": 13020
+    },
+    {
+      "epoch": 0.30686260656587067,
+      "grad_norm": 2.4466586112976074,
+      "learning_rate": 0.00013865573924921106,
+      "loss": 2.1301,
+      "step": 13030
+    },
+    {
+      "epoch": 0.30709811125241393,
+      "grad_norm": 2.5151896476745605,
+      "learning_rate": 0.0001386086383119024,
+      "loss": 2.2717,
+      "step": 13040
+    },
+    {
+      "epoch": 0.3073336159389572,
+      "grad_norm": 2.063638210296631,
+      "learning_rate": 0.00013856153737459376,
+      "loss": 2.0034,
+      "step": 13050
+    },
+    {
+      "epoch": 0.30756912062550046,
+      "grad_norm": 2.0933945178985596,
+      "learning_rate": 0.0001385144364372851,
+      "loss": 2.024,
+      "step": 13060
+    },
+    {
+      "epoch": 0.3078046253120437,
+      "grad_norm": 2.3742902278900146,
+      "learning_rate": 0.00013846733549997646,
+      "loss": 1.9053,
+      "step": 13070
+    },
+    {
+      "epoch": 0.308040129998587,
+      "grad_norm": 2.2515957355499268,
+      "learning_rate": 0.00013842023456266778,
+      "loss": 2.139,
+      "step": 13080
+    },
+    {
+      "epoch": 0.30827563468513025,
+      "grad_norm": 2.533169984817505,
+      "learning_rate": 0.00013837313362535916,
+      "loss": 2.2177,
+      "step": 13090
+    },
+    {
+      "epoch": 0.3085111393716735,
+      "grad_norm": 2.0766286849975586,
+      "learning_rate": 0.00013832603268805048,
+      "loss": 2.0354,
+      "step": 13100
+    },
+    {
+      "epoch": 0.3087466440582168,
+      "grad_norm": 2.1932218074798584,
+      "learning_rate": 0.00013827893175074184,
+      "loss": 2.0149,
+      "step": 13110
+    },
+    {
+      "epoch": 0.30898214874476004,
+      "grad_norm": 2.7338597774505615,
+      "learning_rate": 0.00013823183081343319,
+      "loss": 2.0193,
+      "step": 13120
+    },
+    {
+      "epoch": 0.3092176534313033,
+      "grad_norm": 2.09232497215271,
+      "learning_rate": 0.00013818472987612454,
+      "loss": 2.1508,
+      "step": 13130
+    },
+    {
+      "epoch": 0.30945315811784657,
+      "grad_norm": 2.1449930667877197,
+      "learning_rate": 0.0001381376289388159,
+      "loss": 1.9638,
+      "step": 13140
+    },
+    {
+      "epoch": 0.3096886628043898,
+      "grad_norm": 2.363706350326538,
+      "learning_rate": 0.00013809052800150724,
+      "loss": 2.3092,
+      "step": 13150
+    },
+    {
+      "epoch": 0.30992416749093304,
+      "grad_norm": 2.566270589828491,
+      "learning_rate": 0.00013804342706419859,
+      "loss": 2.0075,
+      "step": 13160
+    },
+    {
+      "epoch": 0.3101596721774763,
+      "grad_norm": 2.4858806133270264,
+      "learning_rate": 0.00013799632612688994,
+      "loss": 2.1862,
+      "step": 13170
+    },
+    {
+      "epoch": 0.3103951768640196,
+      "grad_norm": 2.256286144256592,
+      "learning_rate": 0.0001379492251895813,
+      "loss": 1.9553,
+      "step": 13180
+    },
+    {
+      "epoch": 0.31063068155056284,
+      "grad_norm": 2.1771228313446045,
+      "learning_rate": 0.00013790212425227264,
+      "loss": 2.0341,
+      "step": 13190
+    },
+    {
+      "epoch": 0.3108661862371061,
+      "grad_norm": 2.82309889793396,
+      "learning_rate": 0.000137855023314964,
+      "loss": 2.3205,
+      "step": 13200
+    },
+    {
+      "epoch": 0.31110169092364937,
+      "grad_norm": 2.4577043056488037,
+      "learning_rate": 0.0001378079223776553,
+      "loss": 2.0357,
+      "step": 13210
+    },
+    {
+      "epoch": 0.31133719561019263,
+      "grad_norm": 2.040903091430664,
+      "learning_rate": 0.0001377608214403467,
+      "loss": 2.0322,
+      "step": 13220
+    },
+    {
+      "epoch": 0.3115727002967359,
+      "grad_norm": 2.3443174362182617,
+      "learning_rate": 0.000137713720503038,
+      "loss": 2.0852,
+      "step": 13230
+    },
+    {
+      "epoch": 0.31180820498327916,
+      "grad_norm": 2.266204357147217,
+      "learning_rate": 0.00013766661956572936,
+      "loss": 2.3327,
+      "step": 13240
+    },
+    {
+      "epoch": 0.3120437096698224,
+      "grad_norm": 2.3548195362091064,
+      "learning_rate": 0.0001376195186284207,
+      "loss": 2.1512,
+      "step": 13250
+    },
+    {
+      "epoch": 0.3122792143563657,
+      "grad_norm": 1.9592384099960327,
+      "learning_rate": 0.00013757241769111206,
+      "loss": 2.0161,
+      "step": 13260
+    },
+    {
+      "epoch": 0.31251471904290895,
+      "grad_norm": 2.306349754333496,
+      "learning_rate": 0.0001375253167538034,
+      "loss": 2.1769,
+      "step": 13270
+    },
+    {
+      "epoch": 0.3127502237294522,
+      "grad_norm": 2.76021146774292,
+      "learning_rate": 0.00013747821581649476,
+      "loss": 2.3517,
+      "step": 13280
+    },
+    {
+      "epoch": 0.3129857284159955,
+      "grad_norm": 3.3570475578308105,
+      "learning_rate": 0.00013743111487918609,
+      "loss": 2.1244,
+      "step": 13290
+    },
+    {
+      "epoch": 0.31322123310253874,
+      "grad_norm": 2.5103135108947754,
+      "learning_rate": 0.00013738401394187746,
+      "loss": 2.057,
+      "step": 13300
+    },
+    {
+      "epoch": 0.313456737789082,
+      "grad_norm": 2.2112882137298584,
+      "learning_rate": 0.00013733691300456879,
+      "loss": 1.9779,
+      "step": 13310
+    },
+    {
+      "epoch": 0.31369224247562527,
+      "grad_norm": 2.6416406631469727,
+      "learning_rate": 0.00013728981206726014,
+      "loss": 2.1394,
+      "step": 13320
+    },
+    {
+      "epoch": 0.31392774716216854,
+      "grad_norm": 1.99900484085083,
+      "learning_rate": 0.00013724271112995149,
+      "loss": 2.3333,
+      "step": 13330
+    },
+    {
+      "epoch": 0.3141632518487118,
+      "grad_norm": 2.0924324989318848,
+      "learning_rate": 0.00013719561019264284,
+      "loss": 2.0693,
+      "step": 13340
+    },
+    {
+      "epoch": 0.31439875653525506,
+      "grad_norm": 2.4140758514404297,
+      "learning_rate": 0.0001371485092553342,
+      "loss": 2.2516,
+      "step": 13350
+    },
+    {
+      "epoch": 0.31463426122179833,
+      "grad_norm": 2.1066081523895264,
+      "learning_rate": 0.00013710140831802554,
+      "loss": 2.1283,
+      "step": 13360
+    },
+    {
+      "epoch": 0.3148697659083416,
+      "grad_norm": 1.8612606525421143,
+      "learning_rate": 0.00013705430738071686,
+      "loss": 2.0633,
+      "step": 13370
+    },
+    {
+      "epoch": 0.31510527059488486,
+      "grad_norm": 2.5588574409484863,
+      "learning_rate": 0.00013700720644340824,
+      "loss": 2.1888,
+      "step": 13380
+    },
+    {
+      "epoch": 0.3153407752814281,
+      "grad_norm": 1.7056317329406738,
+      "learning_rate": 0.00013696010550609956,
+      "loss": 1.9666,
+      "step": 13390
+    },
+    {
+      "epoch": 0.3155762799679714,
+      "grad_norm": 1.8611016273498535,
+      "learning_rate": 0.00013691300456879094,
+      "loss": 2.182,
+      "step": 13400
+    },
+    {
+      "epoch": 0.31581178465451465,
+      "grad_norm": 2.0268967151641846,
+      "learning_rate": 0.0001368659036314823,
+      "loss": 1.8033,
+      "step": 13410
+    },
+    {
+      "epoch": 0.3160472893410579,
+      "grad_norm": 1.7649638652801514,
+      "learning_rate": 0.0001368188026941736,
+      "loss": 1.9693,
+      "step": 13420
+    },
+    {
+      "epoch": 0.3162827940276011,
+      "grad_norm": 3.1262307167053223,
+      "learning_rate": 0.000136771701756865,
+      "loss": 2.1137,
+      "step": 13430
+    },
+    {
+      "epoch": 0.3165182987141444,
+      "grad_norm": 2.176802158355713,
+      "learning_rate": 0.0001367246008195563,
+      "loss": 2.0533,
+      "step": 13440
+    },
+    {
+      "epoch": 0.31675380340068765,
+      "grad_norm": 2.7924866676330566,
+      "learning_rate": 0.00013667749988224766,
+      "loss": 1.9473,
+      "step": 13450
+    },
+    {
+      "epoch": 0.3169893080872309,
+      "grad_norm": 2.8177592754364014,
+      "learning_rate": 0.000136630398944939,
+      "loss": 2.0061,
+      "step": 13460
+    },
+    {
+      "epoch": 0.3172248127737742,
+      "grad_norm": 3.690781354904175,
+      "learning_rate": 0.00013658329800763036,
+      "loss": 2.2083,
+      "step": 13470
+    },
+    {
+      "epoch": 0.31746031746031744,
+      "grad_norm": 2.0106966495513916,
+      "learning_rate": 0.0001365361970703217,
+      "loss": 1.9977,
+      "step": 13480
+    },
+    {
+      "epoch": 0.3176958221468607,
+      "grad_norm": 2.4227206707000732,
+      "learning_rate": 0.00013648909613301306,
+      "loss": 2.0599,
+      "step": 13490
+    },
+    {
+      "epoch": 0.31793132683340397,
+      "grad_norm": 2.597426176071167,
+      "learning_rate": 0.00013644199519570439,
+      "loss": 2.0478,
+      "step": 13500
+    },
+    {
+      "epoch": 0.31816683151994724,
+      "grad_norm": 2.364302158355713,
+      "learning_rate": 0.00013639489425839576,
+      "loss": 2.1535,
+      "step": 13510
+    },
+    {
+      "epoch": 0.3184023362064905,
+      "grad_norm": 1.8173555135726929,
+      "learning_rate": 0.0001363477933210871,
+      "loss": 2.0284,
+      "step": 13520
+    },
+    {
+      "epoch": 0.31863784089303376,
+      "grad_norm": 2.0568530559539795,
+      "learning_rate": 0.00013630069238377844,
+      "loss": 1.9601,
+      "step": 13530
+    },
+    {
+      "epoch": 0.318873345579577,
+      "grad_norm": 2.300703525543213,
+      "learning_rate": 0.0001362535914464698,
+      "loss": 2.1593,
+      "step": 13540
+    },
+    {
+      "epoch": 0.3191088502661203,
+      "grad_norm": 2.1626505851745605,
+      "learning_rate": 0.00013620649050916114,
+      "loss": 2.0686,
+      "step": 13550
+    },
+    {
+      "epoch": 0.31934435495266356,
+      "grad_norm": 2.2925024032592773,
+      "learning_rate": 0.0001361593895718525,
+      "loss": 2.0937,
+      "step": 13560
+    },
+    {
+      "epoch": 0.3195798596392068,
+      "grad_norm": 2.871372938156128,
+      "learning_rate": 0.00013611228863454384,
+      "loss": 2.2753,
+      "step": 13570
+    },
+    {
+      "epoch": 0.3198153643257501,
+      "grad_norm": 2.3106420040130615,
+      "learning_rate": 0.0001360651876972352,
+      "loss": 2.2251,
+      "step": 13580
+    },
+    {
+      "epoch": 0.32005086901229335,
+      "grad_norm": 2.192394256591797,
+      "learning_rate": 0.00013601808675992654,
+      "loss": 2.1868,
+      "step": 13590
+    },
+    {
+      "epoch": 0.3202863736988366,
+      "grad_norm": 2.862781286239624,
+      "learning_rate": 0.00013597098582261786,
+      "loss": 2.1841,
+      "step": 13600
+    },
+    {
+      "epoch": 0.3205218783853799,
+      "grad_norm": 2.3184902667999268,
+      "learning_rate": 0.00013592388488530924,
+      "loss": 2.3279,
+      "step": 13610
+    },
+    {
+      "epoch": 0.32075738307192314,
+      "grad_norm": 2.3771584033966064,
+      "learning_rate": 0.00013587678394800056,
+      "loss": 2.2126,
+      "step": 13620
+    },
+    {
+      "epoch": 0.3209928877584664,
+      "grad_norm": 1.6889021396636963,
+      "learning_rate": 0.0001358296830106919,
+      "loss": 2.0039,
+      "step": 13630
+    },
+    {
+      "epoch": 0.32122839244500967,
+      "grad_norm": 2.208984613418579,
+      "learning_rate": 0.00013578258207338326,
+      "loss": 1.9929,
+      "step": 13640
+    },
+    {
+      "epoch": 0.32146389713155293,
+      "grad_norm": 2.5552730560302734,
+      "learning_rate": 0.0001357354811360746,
+      "loss": 2.0772,
+      "step": 13650
+    },
+    {
+      "epoch": 0.3216994018180962,
+      "grad_norm": 2.576536178588867,
+      "learning_rate": 0.00013568838019876596,
+      "loss": 2.0566,
+      "step": 13660
+    },
+    {
+      "epoch": 0.32193490650463946,
+      "grad_norm": 3.376354694366455,
+      "learning_rate": 0.0001356412792614573,
+      "loss": 2.0092,
+      "step": 13670
+    },
+    {
+      "epoch": 0.3221704111911827,
+      "grad_norm": 2.312920331954956,
+      "learning_rate": 0.00013559417832414864,
+      "loss": 2.1,
+      "step": 13680
+    },
+    {
+      "epoch": 0.322405915877726,
+      "grad_norm": 2.1818175315856934,
+      "learning_rate": 0.00013554707738684001,
+      "loss": 1.9808,
+      "step": 13690
+    },
+    {
+      "epoch": 0.32264142056426925,
+      "grad_norm": 2.547168493270874,
+      "learning_rate": 0.00013549997644953136,
+      "loss": 2.1154,
+      "step": 13700
+    },
+    {
+      "epoch": 0.3228769252508125,
+      "grad_norm": 1.7971171140670776,
+      "learning_rate": 0.0001354528755122227,
+      "loss": 2.126,
+      "step": 13710
+    },
+    {
+      "epoch": 0.3231124299373557,
+      "grad_norm": 2.0792088508605957,
+      "learning_rate": 0.00013540577457491406,
+      "loss": 1.9622,
+      "step": 13720
+    },
+    {
+      "epoch": 0.323347934623899,
+      "grad_norm": 1.9572780132293701,
+      "learning_rate": 0.0001353586736376054,
+      "loss": 1.9715,
+      "step": 13730
+    },
+    {
+      "epoch": 0.32358343931044226,
+      "grad_norm": 2.749486207962036,
+      "learning_rate": 0.00013531157270029676,
+      "loss": 2.2712,
+      "step": 13740
+    },
+    {
+      "epoch": 0.3238189439969855,
+      "grad_norm": 2.157297372817993,
+      "learning_rate": 0.0001352644717629881,
+      "loss": 2.2583,
+      "step": 13750
+    },
+    {
+      "epoch": 0.3240544486835288,
+      "grad_norm": 2.2807443141937256,
+      "learning_rate": 0.00013521737082567944,
+      "loss": 2.1891,
+      "step": 13760
+    },
+    {
+      "epoch": 0.32428995337007205,
+      "grad_norm": 1.8038594722747803,
+      "learning_rate": 0.0001351702698883708,
+      "loss": 1.9942,
+      "step": 13770
+    },
+    {
+      "epoch": 0.3245254580566153,
+      "grad_norm": 2.015502691268921,
+      "learning_rate": 0.00013512316895106214,
+      "loss": 1.9229,
+      "step": 13780
+    },
+    {
+      "epoch": 0.3247609627431586,
+      "grad_norm": 2.0219409465789795,
+      "learning_rate": 0.0001350760680137535,
+      "loss": 2.0995,
+      "step": 13790
+    },
+    {
+      "epoch": 0.32499646742970184,
+      "grad_norm": 2.005467176437378,
+      "learning_rate": 0.00013502896707644484,
+      "loss": 2.196,
+      "step": 13800
+    },
+    {
+      "epoch": 0.3252319721162451,
+      "grad_norm": 2.505066394805908,
+      "learning_rate": 0.00013498186613913616,
+      "loss": 2.0781,
+      "step": 13810
+    },
+    {
+      "epoch": 0.32546747680278837,
+      "grad_norm": 2.1414785385131836,
+      "learning_rate": 0.00013493476520182754,
+      "loss": 2.0771,
+      "step": 13820
+    },
+    {
+      "epoch": 0.32570298148933163,
+      "grad_norm": 2.4025304317474365,
+      "learning_rate": 0.00013488766426451886,
+      "loss": 2.2001,
+      "step": 13830
+    },
+    {
+      "epoch": 0.3259384861758749,
+      "grad_norm": 2.1574246883392334,
+      "learning_rate": 0.0001348405633272102,
+      "loss": 2.0484,
+      "step": 13840
+    },
+    {
+      "epoch": 0.32617399086241816,
+      "grad_norm": 2.238400936126709,
+      "learning_rate": 0.00013479346238990156,
+      "loss": 2.0084,
+      "step": 13850
+    },
+    {
+      "epoch": 0.3264094955489614,
+      "grad_norm": 1.7974083423614502,
+      "learning_rate": 0.00013474636145259291,
+      "loss": 2.1113,
+      "step": 13860
+    },
+    {
+      "epoch": 0.3266450002355047,
+      "grad_norm": 2.17846941947937,
+      "learning_rate": 0.00013469926051528426,
+      "loss": 2.3447,
+      "step": 13870
+    },
+    {
+      "epoch": 0.32688050492204795,
+      "grad_norm": 2.0620272159576416,
+      "learning_rate": 0.00013465215957797561,
+      "loss": 2.1059,
+      "step": 13880
+    },
+    {
+      "epoch": 0.3271160096085912,
+      "grad_norm": 2.4582271575927734,
+      "learning_rate": 0.00013460505864066694,
+      "loss": 1.8575,
+      "step": 13890
+    },
+    {
+      "epoch": 0.3273515142951345,
+      "grad_norm": 2.7777035236358643,
+      "learning_rate": 0.00013455795770335831,
+      "loss": 2.1932,
+      "step": 13900
+    },
+    {
+      "epoch": 0.32758701898167775,
+      "grad_norm": 1.6070411205291748,
+      "learning_rate": 0.00013451085676604964,
+      "loss": 1.9579,
+      "step": 13910
+    },
+    {
+      "epoch": 0.327822523668221,
+      "grad_norm": 2.466210126876831,
+      "learning_rate": 0.000134463755828741,
+      "loss": 1.9896,
+      "step": 13920
+    },
+    {
+      "epoch": 0.3280580283547643,
+      "grad_norm": 1.9184893369674683,
+      "learning_rate": 0.00013441665489143234,
+      "loss": 2.2178,
+      "step": 13930
+    },
+    {
+      "epoch": 0.32829353304130754,
+      "grad_norm": 2.1110756397247314,
+      "learning_rate": 0.0001343695539541237,
+      "loss": 2.1961,
+      "step": 13940
+    },
+    {
+      "epoch": 0.3285290377278508,
+      "grad_norm": 2.830181121826172,
+      "learning_rate": 0.00013432245301681504,
+      "loss": 2.1987,
+      "step": 13950
+    },
+    {
+      "epoch": 0.32876454241439407,
+      "grad_norm": 2.1129209995269775,
+      "learning_rate": 0.0001342753520795064,
+      "loss": 2.0927,
+      "step": 13960
+    },
+    {
+      "epoch": 0.32900004710093733,
+      "grad_norm": 2.1881232261657715,
+      "learning_rate": 0.0001342282511421977,
+      "loss": 1.8427,
+      "step": 13970
+    },
+    {
+      "epoch": 0.3292355517874806,
+      "grad_norm": 2.1353397369384766,
+      "learning_rate": 0.0001341811502048891,
+      "loss": 1.9289,
+      "step": 13980
+    },
+    {
+      "epoch": 0.32947105647402386,
+      "grad_norm": 3.527761459350586,
+      "learning_rate": 0.00013413404926758044,
+      "loss": 2.2088,
+      "step": 13990
+    },
+    {
+      "epoch": 0.32970656116056707,
+      "grad_norm": 2.0877022743225098,
+      "learning_rate": 0.0001340869483302718,
+      "loss": 2.1233,
+      "step": 14000
+    },
+    {
+      "epoch": 0.32994206584711033,
+      "grad_norm": 1.560817837715149,
+      "learning_rate": 0.00013403984739296314,
+      "loss": 2.0715,
+      "step": 14010
+    },
+    {
+      "epoch": 0.3301775705336536,
+      "grad_norm": 2.216294765472412,
+      "learning_rate": 0.00013399274645565446,
+      "loss": 2.0866,
+      "step": 14020
+    },
+    {
+      "epoch": 0.33041307522019686,
+      "grad_norm": 2.3411684036254883,
+      "learning_rate": 0.00013394564551834584,
+      "loss": 2.1824,
+      "step": 14030
+    },
+    {
+      "epoch": 0.3306485799067401,
+      "grad_norm": 2.613354206085205,
+      "learning_rate": 0.00013389854458103716,
+      "loss": 2.3151,
+      "step": 14040
+    },
+    {
+      "epoch": 0.3308840845932834,
+      "grad_norm": 2.107166290283203,
+      "learning_rate": 0.00013385144364372851,
+      "loss": 2.091,
+      "step": 14050
+    },
+    {
+      "epoch": 0.33111958927982665,
+      "grad_norm": 2.1655850410461426,
+      "learning_rate": 0.00013380434270641986,
+      "loss": 2.2143,
+      "step": 14060
+    },
+    {
+      "epoch": 0.3313550939663699,
+      "grad_norm": 1.9823925495147705,
+      "learning_rate": 0.00013375724176911121,
+      "loss": 2.1066,
+      "step": 14070
+    },
+    {
+      "epoch": 0.3315905986529132,
+      "grad_norm": 2.26995849609375,
+      "learning_rate": 0.00013371014083180256,
+      "loss": 1.9626,
+      "step": 14080
+    },
+    {
+      "epoch": 0.33182610333945645,
+      "grad_norm": 1.9416958093643188,
+      "learning_rate": 0.00013366303989449392,
+      "loss": 2.2461,
+      "step": 14090
+    },
+    {
+      "epoch": 0.3320616080259997,
+      "grad_norm": 2.685157060623169,
+      "learning_rate": 0.00013361593895718524,
+      "loss": 1.9703,
+      "step": 14100
+    },
+    {
+      "epoch": 0.332297112712543,
+      "grad_norm": 2.3456876277923584,
+      "learning_rate": 0.00013356883801987662,
+      "loss": 2.2025,
+      "step": 14110
+    },
+    {
+      "epoch": 0.33253261739908624,
+      "grad_norm": 2.766294002532959,
+      "learning_rate": 0.00013352173708256794,
+      "loss": 2.2302,
+      "step": 14120
+    },
+    {
+      "epoch": 0.3327681220856295,
+      "grad_norm": 2.4374420642852783,
+      "learning_rate": 0.0001334746361452593,
+      "loss": 2.1317,
+      "step": 14130
+    },
+    {
+      "epoch": 0.33300362677217277,
+      "grad_norm": 1.8526066541671753,
+      "learning_rate": 0.00013342753520795064,
+      "loss": 2.0103,
+      "step": 14140
+    },
+    {
+      "epoch": 0.33323913145871603,
+      "grad_norm": 2.461101770401001,
+      "learning_rate": 0.000133380434270642,
+      "loss": 2.0952,
+      "step": 14150
+    },
+    {
+      "epoch": 0.3334746361452593,
+      "grad_norm": 1.950975775718689,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 2.0181,
+      "step": 14160
+    },
+    {
+      "epoch": 0.33371014083180256,
+      "grad_norm": 2.2419090270996094,
+      "learning_rate": 0.0001332862323960247,
+      "loss": 2.226,
+      "step": 14170
+    },
+    {
+      "epoch": 0.3339456455183458,
+      "grad_norm": 2.077028751373291,
+      "learning_rate": 0.000133239131458716,
+      "loss": 2.0689,
+      "step": 14180
+    },
+    {
+      "epoch": 0.3341811502048891,
+      "grad_norm": 1.9552679061889648,
+      "learning_rate": 0.0001331920305214074,
+      "loss": 2.219,
+      "step": 14190
+    },
+    {
+      "epoch": 0.33441665489143235,
+      "grad_norm": 1.9980392456054688,
+      "learning_rate": 0.00013314492958409871,
+      "loss": 2.1455,
+      "step": 14200
+    },
+    {
+      "epoch": 0.3346521595779756,
+      "grad_norm": 2.1066527366638184,
+      "learning_rate": 0.0001330978286467901,
+      "loss": 2.413,
+      "step": 14210
+    },
+    {
+      "epoch": 0.3348876642645189,
+      "grad_norm": 2.242450475692749,
+      "learning_rate": 0.00013305072770948141,
+      "loss": 2.0,
+      "step": 14220
+    },
+    {
+      "epoch": 0.33512316895106214,
+      "grad_norm": 2.308070659637451,
+      "learning_rate": 0.00013300362677217276,
+      "loss": 1.9766,
+      "step": 14230
+    },
+    {
+      "epoch": 0.3353586736376054,
+      "grad_norm": 3.0701427459716797,
+      "learning_rate": 0.00013295652583486411,
+      "loss": 2.0551,
+      "step": 14240
+    },
+    {
+      "epoch": 0.3355941783241487,
+      "grad_norm": 2.204287052154541,
+      "learning_rate": 0.00013290942489755546,
+      "loss": 2.1141,
+      "step": 14250
+    },
+    {
+      "epoch": 0.33582968301069194,
+      "grad_norm": 2.2963192462921143,
+      "learning_rate": 0.00013286232396024682,
+      "loss": 1.8639,
+      "step": 14260
+    },
+    {
+      "epoch": 0.3360651876972352,
+      "grad_norm": 1.9334232807159424,
+      "learning_rate": 0.00013281522302293817,
+      "loss": 1.9872,
+      "step": 14270
+    },
+    {
+      "epoch": 0.3363006923837784,
+      "grad_norm": 2.3119375705718994,
+      "learning_rate": 0.00013276812208562952,
+      "loss": 2.2071,
+      "step": 14280
+    },
+    {
+      "epoch": 0.3365361970703217,
+      "grad_norm": 1.685998797416687,
+      "learning_rate": 0.00013272102114832087,
+      "loss": 2.2305,
+      "step": 14290
+    },
+    {
+      "epoch": 0.33677170175686494,
+      "grad_norm": 1.8673522472381592,
+      "learning_rate": 0.00013267392021101222,
+      "loss": 2.0198,
+      "step": 14300
+    },
+    {
+      "epoch": 0.3370072064434082,
+      "grad_norm": 2.057691812515259,
+      "learning_rate": 0.00013262681927370354,
+      "loss": 2.1267,
+      "step": 14310
+    },
+    {
+      "epoch": 0.33724271112995147,
+      "grad_norm": 2.4289650917053223,
+      "learning_rate": 0.00013257971833639492,
+      "loss": 2.1753,
+      "step": 14320
+    },
+    {
+      "epoch": 0.33747821581649473,
+      "grad_norm": 2.0615077018737793,
+      "learning_rate": 0.00013253261739908624,
+      "loss": 2.1745,
+      "step": 14330
+    },
+    {
+      "epoch": 0.337713720503038,
+      "grad_norm": 1.9922257661819458,
+      "learning_rate": 0.0001324855164617776,
+      "loss": 2.041,
+      "step": 14340
+    },
+    {
+      "epoch": 0.33794922518958126,
+      "grad_norm": 2.150548219680786,
+      "learning_rate": 0.00013243841552446894,
+      "loss": 2.054,
+      "step": 14350
+    },
+    {
+      "epoch": 0.3381847298761245,
+      "grad_norm": 1.9815877676010132,
+      "learning_rate": 0.0001323913145871603,
+      "loss": 2.2345,
+      "step": 14360
+    },
+    {
+      "epoch": 0.3384202345626678,
+      "grad_norm": 3.1625378131866455,
+      "learning_rate": 0.00013234421364985164,
+      "loss": 2.0449,
+      "step": 14370
+    },
+    {
+      "epoch": 0.33865573924921105,
+      "grad_norm": 2.5849571228027344,
+      "learning_rate": 0.000132297112712543,
+      "loss": 2.1497,
+      "step": 14380
+    },
+    {
+      "epoch": 0.3388912439357543,
+      "grad_norm": 3.0751826763153076,
+      "learning_rate": 0.00013225001177523434,
+      "loss": 2.0963,
+      "step": 14390
+    },
+    {
+      "epoch": 0.3391267486222976,
+      "grad_norm": 1.572991967201233,
+      "learning_rate": 0.0001322029108379257,
+      "loss": 1.9016,
+      "step": 14400
+    },
+    {
+      "epoch": 0.33936225330884084,
+      "grad_norm": 2.5276403427124023,
+      "learning_rate": 0.00013215580990061701,
+      "loss": 2.3332,
+      "step": 14410
+    },
+    {
+      "epoch": 0.3395977579953841,
+      "grad_norm": 1.7656186819076538,
+      "learning_rate": 0.0001321087089633084,
+      "loss": 2.0629,
+      "step": 14420
+    },
+    {
+      "epoch": 0.3398332626819274,
+      "grad_norm": 1.8778975009918213,
+      "learning_rate": 0.00013206160802599972,
+      "loss": 2.2139,
+      "step": 14430
+    },
+    {
+      "epoch": 0.34006876736847064,
+      "grad_norm": 2.696181535720825,
+      "learning_rate": 0.00013201450708869107,
+      "loss": 1.9954,
+      "step": 14440
+    },
+    {
+      "epoch": 0.3403042720550139,
+      "grad_norm": 1.6417945623397827,
+      "learning_rate": 0.00013196740615138242,
+      "loss": 1.9359,
+      "step": 14450
+    },
+    {
+      "epoch": 0.34053977674155717,
+      "grad_norm": 2.1908366680145264,
+      "learning_rate": 0.00013192030521407377,
+      "loss": 1.9072,
+      "step": 14460
+    },
+    {
+      "epoch": 0.34077528142810043,
+      "grad_norm": 2.7889277935028076,
+      "learning_rate": 0.00013187320427676512,
+      "loss": 2.0709,
+      "step": 14470
+    },
+    {
+      "epoch": 0.3410107861146437,
+      "grad_norm": 2.474809408187866,
+      "learning_rate": 0.00013182610333945647,
+      "loss": 2.155,
+      "step": 14480
+    },
+    {
+      "epoch": 0.34124629080118696,
+      "grad_norm": 2.2163426876068115,
+      "learning_rate": 0.0001317790024021478,
+      "loss": 2.1198,
+      "step": 14490
+    },
+    {
+      "epoch": 0.3414817954877302,
+      "grad_norm": 2.4669189453125,
+      "learning_rate": 0.00013173190146483917,
+      "loss": 2.1101,
+      "step": 14500
+    },
+    {
+      "epoch": 0.3417173001742735,
+      "grad_norm": 2.4319045543670654,
+      "learning_rate": 0.0001316848005275305,
+      "loss": 2.208,
+      "step": 14510
+    },
+    {
+      "epoch": 0.34195280486081675,
+      "grad_norm": 2.244265556335449,
+      "learning_rate": 0.00013163769959022184,
+      "loss": 2.0715,
+      "step": 14520
+    },
+    {
+      "epoch": 0.34218830954736,
+      "grad_norm": 2.1139960289001465,
+      "learning_rate": 0.0001315905986529132,
+      "loss": 1.9087,
+      "step": 14530
+    },
+    {
+      "epoch": 0.3424238142339033,
+      "grad_norm": 2.030888080596924,
+      "learning_rate": 0.00013154349771560454,
+      "loss": 2.0678,
+      "step": 14540
+    },
+    {
+      "epoch": 0.34265931892044654,
+      "grad_norm": 3.3537919521331787,
+      "learning_rate": 0.00013149639677829592,
+      "loss": 2.1527,
+      "step": 14550
+    },
+    {
+      "epoch": 0.34289482360698975,
+      "grad_norm": 2.137396812438965,
+      "learning_rate": 0.00013144929584098724,
+      "loss": 2.2981,
+      "step": 14560
+    },
+    {
+      "epoch": 0.343130328293533,
+      "grad_norm": 3.0466394424438477,
+      "learning_rate": 0.0001314021949036786,
+      "loss": 2.1918,
+      "step": 14570
+    },
+    {
+      "epoch": 0.3433658329800763,
+      "grad_norm": 2.6642491817474365,
+      "learning_rate": 0.00013135509396636994,
+      "loss": 2.2627,
+      "step": 14580
+    },
+    {
+      "epoch": 0.34360133766661954,
+      "grad_norm": 1.9378482103347778,
+      "learning_rate": 0.0001313079930290613,
+      "loss": 2.0302,
+      "step": 14590
+    },
+    {
+      "epoch": 0.3438368423531628,
+      "grad_norm": 2.130795955657959,
+      "learning_rate": 0.00013126089209175264,
+      "loss": 2.0724,
+      "step": 14600
+    },
+    {
+      "epoch": 0.3440723470397061,
+      "grad_norm": 3.243426561355591,
+      "learning_rate": 0.000131213791154444,
+      "loss": 2.1017,
+      "step": 14610
+    },
+    {
+      "epoch": 0.34430785172624934,
+      "grad_norm": 2.094165086746216,
+      "learning_rate": 0.00013116669021713532,
+      "loss": 2.2293,
+      "step": 14620
+    },
+    {
+      "epoch": 0.3445433564127926,
+      "grad_norm": 2.4611902236938477,
+      "learning_rate": 0.0001311195892798267,
+      "loss": 2.0473,
+      "step": 14630
+    },
+    {
+      "epoch": 0.34477886109933586,
+      "grad_norm": 2.694459915161133,
+      "learning_rate": 0.00013107248834251802,
+      "loss": 2.0456,
+      "step": 14640
+    },
+    {
+      "epoch": 0.34501436578587913,
+      "grad_norm": 1.9303042888641357,
+      "learning_rate": 0.00013102538740520937,
+      "loss": 2.0716,
+      "step": 14650
+    },
+    {
+      "epoch": 0.3452498704724224,
+      "grad_norm": 2.9043257236480713,
+      "learning_rate": 0.00013097828646790072,
+      "loss": 2.1234,
+      "step": 14660
+    },
+    {
+      "epoch": 0.34548537515896566,
+      "grad_norm": 2.1518394947052,
+      "learning_rate": 0.00013093118553059207,
+      "loss": 2.1043,
+      "step": 14670
+    },
+    {
+      "epoch": 0.3457208798455089,
+      "grad_norm": 2.1524922847747803,
+      "learning_rate": 0.00013088408459328342,
+      "loss": 2.2017,
+      "step": 14680
+    },
+    {
+      "epoch": 0.3459563845320522,
+      "grad_norm": 3.270179510116577,
+      "learning_rate": 0.00013083698365597477,
+      "loss": 2.0904,
+      "step": 14690
+    },
+    {
+      "epoch": 0.34619188921859545,
+      "grad_norm": 2.349952220916748,
+      "learning_rate": 0.0001307898827186661,
+      "loss": 1.9607,
+      "step": 14700
+    },
+    {
+      "epoch": 0.3464273939051387,
+      "grad_norm": 2.087453842163086,
+      "learning_rate": 0.00013074278178135747,
+      "loss": 2.0567,
+      "step": 14710
+    },
+    {
+      "epoch": 0.346662898591682,
+      "grad_norm": 1.739876389503479,
+      "learning_rate": 0.0001306956808440488,
+      "loss": 1.9899,
+      "step": 14720
+    },
+    {
+      "epoch": 0.34689840327822524,
+      "grad_norm": 2.5640616416931152,
+      "learning_rate": 0.00013064857990674014,
+      "loss": 2.126,
+      "step": 14730
+    },
+    {
+      "epoch": 0.3471339079647685,
+      "grad_norm": 2.1961350440979004,
+      "learning_rate": 0.0001306014789694315,
+      "loss": 2.2033,
+      "step": 14740
+    },
+    {
+      "epoch": 0.34736941265131177,
+      "grad_norm": 2.4550042152404785,
+      "learning_rate": 0.00013055437803212284,
+      "loss": 2.2094,
+      "step": 14750
+    },
+    {
+      "epoch": 0.34760491733785503,
+      "grad_norm": 2.9043076038360596,
+      "learning_rate": 0.0001305072770948142,
+      "loss": 2.1208,
+      "step": 14760
+    },
+    {
+      "epoch": 0.3478404220243983,
+      "grad_norm": 2.2261176109313965,
+      "learning_rate": 0.00013046017615750554,
+      "loss": 2.3374,
+      "step": 14770
+    },
+    {
+      "epoch": 0.34807592671094156,
+      "grad_norm": 3.6236088275909424,
+      "learning_rate": 0.00013041307522019687,
+      "loss": 2.1553,
+      "step": 14780
+    },
+    {
+      "epoch": 0.3483114313974848,
+      "grad_norm": 1.7490025758743286,
+      "learning_rate": 0.00013036597428288824,
+      "loss": 2.0613,
+      "step": 14790
+    },
+    {
+      "epoch": 0.3485469360840281,
+      "grad_norm": 1.8174374103546143,
+      "learning_rate": 0.00013031887334557957,
+      "loss": 1.9779,
+      "step": 14800
+    },
+    {
+      "epoch": 0.34878244077057136,
+      "grad_norm": 2.0383830070495605,
+      "learning_rate": 0.00013027177240827094,
+      "loss": 2.0438,
+      "step": 14810
+    },
+    {
+      "epoch": 0.3490179454571146,
+      "grad_norm": 2.6745285987854004,
+      "learning_rate": 0.0001302246714709623,
+      "loss": 2.0689,
+      "step": 14820
+    },
+    {
+      "epoch": 0.3492534501436579,
+      "grad_norm": 2.264727830886841,
+      "learning_rate": 0.00013017757053365362,
+      "loss": 2.3443,
+      "step": 14830
+    },
+    {
+      "epoch": 0.34948895483020115,
+      "grad_norm": 1.7950100898742676,
+      "learning_rate": 0.000130130469596345,
+      "loss": 2.101,
+      "step": 14840
+    },
+    {
+      "epoch": 0.34972445951674436,
+      "grad_norm": 2.4111218452453613,
+      "learning_rate": 0.00013008336865903632,
+      "loss": 1.9223,
+      "step": 14850
+    },
+    {
+      "epoch": 0.3499599642032876,
+      "grad_norm": 2.078031539916992,
+      "learning_rate": 0.00013003626772172767,
+      "loss": 2.1652,
+      "step": 14860
+    },
+    {
+      "epoch": 0.3501954688898309,
+      "grad_norm": 2.400324583053589,
+      "learning_rate": 0.00012998916678441902,
+      "loss": 1.9705,
+      "step": 14870
+    },
+    {
+      "epoch": 0.35043097357637415,
+      "grad_norm": 1.9067310094833374,
+      "learning_rate": 0.00012994206584711037,
+      "loss": 2.0825,
+      "step": 14880
+    },
+    {
+      "epoch": 0.3506664782629174,
+      "grad_norm": 2.236149549484253,
+      "learning_rate": 0.00012989496490980172,
+      "loss": 2.2749,
+      "step": 14890
+    },
+    {
+      "epoch": 0.3509019829494607,
+      "grad_norm": 1.9904561042785645,
+      "learning_rate": 0.00012984786397249307,
+      "loss": 2.0971,
+      "step": 14900
+    },
+    {
+      "epoch": 0.35113748763600394,
+      "grad_norm": 2.2129764556884766,
+      "learning_rate": 0.0001298007630351844,
+      "loss": 2.2559,
+      "step": 14910
+    },
+    {
+      "epoch": 0.3513729923225472,
+      "grad_norm": 2.020630121231079,
+      "learning_rate": 0.00012975366209787577,
+      "loss": 1.968,
+      "step": 14920
+    },
+    {
+      "epoch": 0.35160849700909047,
+      "grad_norm": 3.8411476612091064,
+      "learning_rate": 0.0001297065611605671,
+      "loss": 2.0724,
+      "step": 14930
+    },
+    {
+      "epoch": 0.35184400169563373,
+      "grad_norm": 4.267243385314941,
+      "learning_rate": 0.00012965946022325844,
+      "loss": 1.9732,
+      "step": 14940
+    },
+    {
+      "epoch": 0.352079506382177,
+      "grad_norm": 2.2827847003936768,
+      "learning_rate": 0.0001296123592859498,
+      "loss": 2.3295,
+      "step": 14950
+    },
+    {
+      "epoch": 0.35231501106872026,
+      "grad_norm": 2.276663064956665,
+      "learning_rate": 0.00012956525834864114,
+      "loss": 2.2312,
+      "step": 14960
+    },
+    {
+      "epoch": 0.3525505157552635,
+      "grad_norm": 2.2448136806488037,
+      "learning_rate": 0.0001295181574113325,
+      "loss": 2.0823,
+      "step": 14970
+    },
+    {
+      "epoch": 0.3527860204418068,
+      "grad_norm": 2.307004690170288,
+      "learning_rate": 0.00012947105647402384,
+      "loss": 1.9903,
+      "step": 14980
+    },
+    {
+      "epoch": 0.35302152512835006,
+      "grad_norm": 2.2261812686920166,
+      "learning_rate": 0.0001294239555367152,
+      "loss": 2.0342,
+      "step": 14990
+    },
+    {
+      "epoch": 0.3532570298148933,
+      "grad_norm": 2.9633865356445312,
+      "learning_rate": 0.00012937685459940654,
+      "loss": 1.9893,
+      "step": 15000
+    },
+    {
+      "epoch": 0.3534925345014366,
+      "grad_norm": 2.297091007232666,
+      "learning_rate": 0.00012932975366209787,
+      "loss": 2.258,
+      "step": 15010
+    },
+    {
+      "epoch": 0.35372803918797985,
+      "grad_norm": 3.055051803588867,
+      "learning_rate": 0.00012928265272478924,
+      "loss": 2.0553,
+      "step": 15020
+    },
+    {
+      "epoch": 0.3539635438745231,
+      "grad_norm": 1.8271070718765259,
+      "learning_rate": 0.00012923555178748057,
+      "loss": 2.1569,
+      "step": 15030
+    },
+    {
+      "epoch": 0.3541990485610664,
+      "grad_norm": 2.4504764080047607,
+      "learning_rate": 0.00012918845085017192,
+      "loss": 2.1445,
+      "step": 15040
+    },
+    {
+      "epoch": 0.35443455324760964,
+      "grad_norm": 3.338831901550293,
+      "learning_rate": 0.00012914134991286327,
+      "loss": 2.151,
+      "step": 15050
+    },
+    {
+      "epoch": 0.3546700579341529,
+      "grad_norm": 1.9933847188949585,
+      "learning_rate": 0.00012909424897555462,
+      "loss": 1.9229,
+      "step": 15060
+    },
+    {
+      "epoch": 0.35490556262069617,
+      "grad_norm": 1.8507821559906006,
+      "learning_rate": 0.00012904714803824597,
+      "loss": 1.8161,
+      "step": 15070
+    },
+    {
+      "epoch": 0.35514106730723943,
+      "grad_norm": 2.110309362411499,
+      "learning_rate": 0.00012900004710093732,
+      "loss": 2.107,
+      "step": 15080
+    },
+    {
+      "epoch": 0.3553765719937827,
+      "grad_norm": 2.447514057159424,
+      "learning_rate": 0.00012895294616362864,
+      "loss": 2.0256,
+      "step": 15090
+    },
+    {
+      "epoch": 0.35561207668032596,
+      "grad_norm": 2.630833864212036,
+      "learning_rate": 0.00012890584522632002,
+      "loss": 2.2369,
+      "step": 15100
+    },
+    {
+      "epoch": 0.3558475813668692,
+      "grad_norm": 2.210677146911621,
+      "learning_rate": 0.00012885874428901137,
+      "loss": 2.3155,
+      "step": 15110
+    },
+    {
+      "epoch": 0.3560830860534125,
+      "grad_norm": 2.5040535926818848,
+      "learning_rate": 0.0001288116433517027,
+      "loss": 2.1744,
+      "step": 15120
+    },
+    {
+      "epoch": 0.3563185907399557,
+      "grad_norm": 2.41166615486145,
+      "learning_rate": 0.00012876454241439407,
+      "loss": 1.9641,
+      "step": 15130
+    },
+    {
+      "epoch": 0.35655409542649896,
+      "grad_norm": 2.1737725734710693,
+      "learning_rate": 0.0001287174414770854,
+      "loss": 2.0927,
+      "step": 15140
+    },
+    {
+      "epoch": 0.3567896001130422,
+      "grad_norm": 2.3262150287628174,
+      "learning_rate": 0.00012867034053977677,
+      "loss": 1.8485,
+      "step": 15150
+    },
+    {
+      "epoch": 0.3570251047995855,
+      "grad_norm": 2.214301586151123,
+      "learning_rate": 0.0001286232396024681,
+      "loss": 2.0541,
+      "step": 15160
+    },
+    {
+      "epoch": 0.35726060948612876,
+      "grad_norm": 3.0727179050445557,
+      "learning_rate": 0.0001285808487588903,
+      "loss": 2.2335,
+      "step": 15170
+    },
+    {
+      "epoch": 0.357496114172672,
+      "grad_norm": 1.55754816532135,
+      "learning_rate": 0.00012853374782158166,
+      "loss": 1.7324,
+      "step": 15180
+    },
+    {
+      "epoch": 0.3577316188592153,
+      "grad_norm": 2.219801425933838,
+      "learning_rate": 0.000128486646884273,
+      "loss": 2.1554,
+      "step": 15190
+    },
+    {
+      "epoch": 0.35796712354575855,
+      "grad_norm": 2.0826430320739746,
+      "learning_rate": 0.00012843954594696433,
+      "loss": 2.0151,
+      "step": 15200
+    },
+    {
+      "epoch": 0.3582026282323018,
+      "grad_norm": 2.083486557006836,
+      "learning_rate": 0.0001283924450096557,
+      "loss": 2.1411,
+      "step": 15210
+    },
+    {
+      "epoch": 0.3584381329188451,
+      "grad_norm": 2.3072783946990967,
+      "learning_rate": 0.00012834534407234704,
+      "loss": 2.006,
+      "step": 15220
+    },
+    {
+      "epoch": 0.35867363760538834,
+      "grad_norm": 2.440856456756592,
+      "learning_rate": 0.00012829824313503839,
+      "loss": 2.0854,
+      "step": 15230
+    },
+    {
+      "epoch": 0.3589091422919316,
+      "grad_norm": 2.845301866531372,
+      "learning_rate": 0.00012825114219772974,
+      "loss": 2.3434,
+      "step": 15240
+    },
+    {
+      "epoch": 0.35914464697847487,
+      "grad_norm": 2.953895330429077,
+      "learning_rate": 0.00012820404126042109,
+      "loss": 2.0988,
+      "step": 15250
+    },
+    {
+      "epoch": 0.35938015166501813,
+      "grad_norm": 2.4946932792663574,
+      "learning_rate": 0.00012815694032311244,
+      "loss": 2.0624,
+      "step": 15260
+    },
+    {
+      "epoch": 0.3596156563515614,
+      "grad_norm": 2.2791571617126465,
+      "learning_rate": 0.00012810983938580379,
+      "loss": 2.0612,
+      "step": 15270
+    },
+    {
+      "epoch": 0.35985116103810466,
+      "grad_norm": 2.0594568252563477,
+      "learning_rate": 0.00012806273844849514,
+      "loss": 2.2905,
+      "step": 15280
+    },
+    {
+      "epoch": 0.3600866657246479,
+      "grad_norm": 1.937827467918396,
+      "learning_rate": 0.0001280156375111865,
+      "loss": 2.2477,
+      "step": 15290
+    },
+    {
+      "epoch": 0.3603221704111912,
+      "grad_norm": 2.1612651348114014,
+      "learning_rate": 0.0001279685365738778,
+      "loss": 2.064,
+      "step": 15300
+    },
+    {
+      "epoch": 0.36055767509773445,
+      "grad_norm": 2.1727564334869385,
+      "learning_rate": 0.0001279214356365692,
+      "loss": 2.1897,
+      "step": 15310
+    },
+    {
+      "epoch": 0.3607931797842777,
+      "grad_norm": 1.9623788595199585,
+      "learning_rate": 0.0001278743346992605,
+      "loss": 2.0795,
+      "step": 15320
+    },
+    {
+      "epoch": 0.361028684470821,
+      "grad_norm": 2.5158348083496094,
+      "learning_rate": 0.00012782723376195186,
+      "loss": 2.3262,
+      "step": 15330
+    },
+    {
+      "epoch": 0.36126418915736425,
+      "grad_norm": 1.673125982284546,
+      "learning_rate": 0.0001277801328246432,
+      "loss": 2.0423,
+      "step": 15340
+    },
+    {
+      "epoch": 0.3614996938439075,
+      "grad_norm": 2.2631309032440186,
+      "learning_rate": 0.00012773303188733456,
+      "loss": 1.9951,
+      "step": 15350
+    },
+    {
+      "epoch": 0.3617351985304508,
+      "grad_norm": 2.103239059448242,
+      "learning_rate": 0.0001276859309500259,
+      "loss": 2.1707,
+      "step": 15360
+    },
+    {
+      "epoch": 0.36197070321699404,
+      "grad_norm": 2.2068543434143066,
+      "learning_rate": 0.00012763883001271726,
+      "loss": 2.0808,
+      "step": 15370
+    },
+    {
+      "epoch": 0.3622062079035373,
+      "grad_norm": 2.3455021381378174,
+      "learning_rate": 0.0001275917290754086,
+      "loss": 2.1042,
+      "step": 15380
+    },
+    {
+      "epoch": 0.36244171259008057,
+      "grad_norm": 2.655992031097412,
+      "learning_rate": 0.00012754462813809996,
+      "loss": 2.165,
+      "step": 15390
+    },
+    {
+      "epoch": 0.36267721727662383,
+      "grad_norm": 2.5498814582824707,
+      "learning_rate": 0.0001274975272007913,
+      "loss": 2.251,
+      "step": 15400
+    },
+    {
+      "epoch": 0.36291272196316704,
+      "grad_norm": 2.270501136779785,
+      "learning_rate": 0.00012745042626348264,
+      "loss": 1.9743,
+      "step": 15410
+    },
+    {
+      "epoch": 0.3631482266497103,
+      "grad_norm": 1.8444944620132446,
+      "learning_rate": 0.000127403325326174,
+      "loss": 2.2257,
+      "step": 15420
+    },
+    {
+      "epoch": 0.36338373133625357,
+      "grad_norm": 3.1018927097320557,
+      "learning_rate": 0.00012735622438886534,
+      "loss": 2.0163,
+      "step": 15430
+    },
+    {
+      "epoch": 0.36361923602279683,
+      "grad_norm": 3.2892568111419678,
+      "learning_rate": 0.0001273091234515567,
+      "loss": 2.0607,
+      "step": 15440
+    },
+    {
+      "epoch": 0.3638547407093401,
+      "grad_norm": 2.4903275966644287,
+      "learning_rate": 0.00012726202251424804,
+      "loss": 2.0741,
+      "step": 15450
+    },
+    {
+      "epoch": 0.36409024539588336,
+      "grad_norm": 2.295133590698242,
+      "learning_rate": 0.0001272149215769394,
+      "loss": 2.3047,
+      "step": 15460
+    },
+    {
+      "epoch": 0.3643257500824266,
+      "grad_norm": 1.8592169284820557,
+      "learning_rate": 0.00012716782063963074,
+      "loss": 2.1098,
+      "step": 15470
+    },
+    {
+      "epoch": 0.3645612547689699,
+      "grad_norm": 1.8355640172958374,
+      "learning_rate": 0.0001271207197023221,
+      "loss": 2.0824,
+      "step": 15480
+    },
+    {
+      "epoch": 0.36479675945551315,
+      "grad_norm": 2.295816421508789,
+      "learning_rate": 0.00012707361876501344,
+      "loss": 2.1756,
+      "step": 15490
+    },
+    {
+      "epoch": 0.3650322641420564,
+      "grad_norm": 2.0585267543792725,
+      "learning_rate": 0.0001270265178277048,
+      "loss": 2.3163,
+      "step": 15500
+    },
+    {
+      "epoch": 0.3652677688285997,
+      "grad_norm": 1.753099799156189,
+      "learning_rate": 0.0001269794168903961,
+      "loss": 2.1565,
+      "step": 15510
+    },
+    {
+      "epoch": 0.36550327351514295,
+      "grad_norm": 2.6697375774383545,
+      "learning_rate": 0.0001269323159530875,
+      "loss": 2.2484,
+      "step": 15520
+    },
+    {
+      "epoch": 0.3657387782016862,
+      "grad_norm": 2.1869733333587646,
+      "learning_rate": 0.0001268852150157788,
+      "loss": 2.1681,
+      "step": 15530
+    },
+    {
+      "epoch": 0.3659742828882295,
+      "grad_norm": 2.7051522731781006,
+      "learning_rate": 0.00012683811407847016,
+      "loss": 2.0296,
+      "step": 15540
+    },
+    {
+      "epoch": 0.36620978757477274,
+      "grad_norm": 2.1270742416381836,
+      "learning_rate": 0.0001267910131411615,
+      "loss": 1.9708,
+      "step": 15550
+    },
+    {
+      "epoch": 0.366445292261316,
+      "grad_norm": 2.645512342453003,
+      "learning_rate": 0.00012674391220385286,
+      "loss": 2.1499,
+      "step": 15560
+    },
+    {
+      "epoch": 0.36668079694785927,
+      "grad_norm": 2.9818031787872314,
+      "learning_rate": 0.0001266968112665442,
+      "loss": 2.0822,
+      "step": 15570
+    },
+    {
+      "epoch": 0.36691630163440253,
+      "grad_norm": 3.0397183895111084,
+      "learning_rate": 0.00012664971032923556,
+      "loss": 2.1616,
+      "step": 15580
+    },
+    {
+      "epoch": 0.3671518063209458,
+      "grad_norm": 2.2161061763763428,
+      "learning_rate": 0.00012660260939192689,
+      "loss": 1.9094,
+      "step": 15590
+    },
+    {
+      "epoch": 0.36738731100748906,
+      "grad_norm": 1.962997555732727,
+      "learning_rate": 0.00012655550845461826,
+      "loss": 2.0234,
+      "step": 15600
+    },
+    {
+      "epoch": 0.3676228156940323,
+      "grad_norm": 2.1313388347625732,
+      "learning_rate": 0.00012650840751730959,
+      "loss": 2.04,
+      "step": 15610
+    },
+    {
+      "epoch": 0.3678583203805756,
+      "grad_norm": 2.517944574356079,
+      "learning_rate": 0.00012646130658000094,
+      "loss": 2.0798,
+      "step": 15620
+    },
+    {
+      "epoch": 0.36809382506711885,
+      "grad_norm": 2.792917251586914,
+      "learning_rate": 0.0001264142056426923,
+      "loss": 1.9457,
+      "step": 15630
+    },
+    {
+      "epoch": 0.3683293297536621,
+      "grad_norm": 2.193131685256958,
+      "learning_rate": 0.00012636710470538364,
+      "loss": 2.1932,
+      "step": 15640
+    },
+    {
+      "epoch": 0.3685648344402054,
+      "grad_norm": 2.234732151031494,
+      "learning_rate": 0.00012632000376807501,
+      "loss": 2.1695,
+      "step": 15650
+    },
+    {
+      "epoch": 0.36880033912674864,
+      "grad_norm": 2.553999900817871,
+      "learning_rate": 0.00012627290283076634,
+      "loss": 2.1587,
+      "step": 15660
+    },
+    {
+      "epoch": 0.3690358438132919,
+      "grad_norm": 2.0366036891937256,
+      "learning_rate": 0.0001262258018934577,
+      "loss": 2.1545,
+      "step": 15670
+    },
+    {
+      "epoch": 0.3692713484998352,
+      "grad_norm": 1.8725734949111938,
+      "learning_rate": 0.00012617870095614904,
+      "loss": 2.1266,
+      "step": 15680
+    },
+    {
+      "epoch": 0.36950685318637844,
+      "grad_norm": 2.4001288414001465,
+      "learning_rate": 0.0001261316000188404,
+      "loss": 2.1415,
+      "step": 15690
+    },
+    {
+      "epoch": 0.36974235787292165,
+      "grad_norm": 2.269136428833008,
+      "learning_rate": 0.00012608449908153174,
+      "loss": 1.9927,
+      "step": 15700
+    },
+    {
+      "epoch": 0.3699778625594649,
+      "grad_norm": 2.2865076065063477,
+      "learning_rate": 0.0001260373981442231,
+      "loss": 2.3061,
+      "step": 15710
+    },
+    {
+      "epoch": 0.3702133672460082,
+      "grad_norm": 2.2969698905944824,
+      "learning_rate": 0.0001259902972069144,
+      "loss": 2.1753,
+      "step": 15720
+    },
+    {
+      "epoch": 0.37044887193255144,
+      "grad_norm": 2.6389291286468506,
+      "learning_rate": 0.0001259431962696058,
+      "loss": 1.9136,
+      "step": 15730
+    },
+    {
+      "epoch": 0.3706843766190947,
+      "grad_norm": 3.111884355545044,
+      "learning_rate": 0.0001258960953322971,
+      "loss": 1.9592,
+      "step": 15740
+    },
+    {
+      "epoch": 0.37091988130563797,
+      "grad_norm": 1.8911104202270508,
+      "learning_rate": 0.00012584899439498846,
+      "loss": 1.9782,
+      "step": 15750
+    },
+    {
+      "epoch": 0.37115538599218123,
+      "grad_norm": 2.410327672958374,
+      "learning_rate": 0.0001258018934576798,
+      "loss": 2.1944,
+      "step": 15760
+    },
+    {
+      "epoch": 0.3713908906787245,
+      "grad_norm": 2.6267807483673096,
+      "learning_rate": 0.00012575479252037116,
+      "loss": 2.1023,
+      "step": 15770
+    },
+    {
+      "epoch": 0.37162639536526776,
+      "grad_norm": 2.455199718475342,
+      "learning_rate": 0.0001257076915830625,
+      "loss": 2.2238,
+      "step": 15780
+    },
+    {
+      "epoch": 0.371861900051811,
+      "grad_norm": 2.131810188293457,
+      "learning_rate": 0.00012566059064575386,
+      "loss": 2.1704,
+      "step": 15790
+    },
+    {
+      "epoch": 0.3720974047383543,
+      "grad_norm": 2.0262787342071533,
+      "learning_rate": 0.0001256134897084452,
+      "loss": 2.0336,
+      "step": 15800
+    },
+    {
+      "epoch": 0.37233290942489755,
+      "grad_norm": 2.7119791507720947,
+      "learning_rate": 0.00012556638877113656,
+      "loss": 2.0177,
+      "step": 15810
+    },
+    {
+      "epoch": 0.3725684141114408,
+      "grad_norm": 2.81471586227417,
+      "learning_rate": 0.0001255192878338279,
+      "loss": 1.9555,
+      "step": 15820
+    },
+    {
+      "epoch": 0.3728039187979841,
+      "grad_norm": 1.8543721437454224,
+      "learning_rate": 0.00012547218689651924,
+      "loss": 2.1615,
+      "step": 15830
+    },
+    {
+      "epoch": 0.37303942348452734,
+      "grad_norm": 2.129915952682495,
+      "learning_rate": 0.0001254250859592106,
+      "loss": 2.0439,
+      "step": 15840
+    },
+    {
+      "epoch": 0.3732749281710706,
+      "grad_norm": 1.9343938827514648,
+      "learning_rate": 0.00012537798502190194,
+      "loss": 1.944,
+      "step": 15850
+    },
+    {
+      "epoch": 0.37351043285761387,
+      "grad_norm": 2.403035879135132,
+      "learning_rate": 0.0001253308840845933,
+      "loss": 2.0897,
+      "step": 15860
+    },
+    {
+      "epoch": 0.37374593754415714,
+      "grad_norm": 1.8205158710479736,
+      "learning_rate": 0.00012528378314728464,
+      "loss": 2.1931,
+      "step": 15870
+    },
+    {
+      "epoch": 0.3739814422307004,
+      "grad_norm": 2.119952917098999,
+      "learning_rate": 0.000125236682209976,
+      "loss": 2.254,
+      "step": 15880
+    },
+    {
+      "epoch": 0.37421694691724366,
+      "grad_norm": 2.8081493377685547,
+      "learning_rate": 0.00012518958127266734,
+      "loss": 2.0929,
+      "step": 15890
+    },
+    {
+      "epoch": 0.37445245160378693,
+      "grad_norm": 1.801512598991394,
+      "learning_rate": 0.00012514248033535866,
+      "loss": 2.1442,
+      "step": 15900
+    },
+    {
+      "epoch": 0.3746879562903302,
+      "grad_norm": 2.782501697540283,
+      "learning_rate": 0.00012509537939805004,
+      "loss": 2.1213,
+      "step": 15910
+    },
+    {
+      "epoch": 0.37492346097687346,
+      "grad_norm": 2.08080792427063,
+      "learning_rate": 0.00012504827846074136,
+      "loss": 2.0139,
+      "step": 15920
+    },
+    {
+      "epoch": 0.3751589656634167,
+      "grad_norm": 2.6191723346710205,
+      "learning_rate": 0.0001250011775234327,
+      "loss": 2.1154,
+      "step": 15930
+    },
+    {
+      "epoch": 0.37539447034996,
+      "grad_norm": 2.5812697410583496,
+      "learning_rate": 0.0001249540765861241,
+      "loss": 1.9264,
+      "step": 15940
+    },
+    {
+      "epoch": 0.37562997503650325,
+      "grad_norm": 1.9726451635360718,
+      "learning_rate": 0.0001249069756488154,
+      "loss": 2.1646,
+      "step": 15950
+    },
+    {
+      "epoch": 0.3758654797230465,
+      "grad_norm": 1.8344637155532837,
+      "learning_rate": 0.00012485987471150676,
+      "loss": 2.1906,
+      "step": 15960
+    },
+    {
+      "epoch": 0.3761009844095898,
+      "grad_norm": 2.092193365097046,
+      "learning_rate": 0.00012481277377419811,
+      "loss": 2.2604,
+      "step": 15970
+    },
+    {
+      "epoch": 0.376336489096133,
+      "grad_norm": 1.8692036867141724,
+      "learning_rate": 0.00012476567283688946,
+      "loss": 1.9886,
+      "step": 15980
+    },
+    {
+      "epoch": 0.37657199378267625,
+      "grad_norm": 1.7763489484786987,
+      "learning_rate": 0.00012471857189958081,
+      "loss": 2.0102,
+      "step": 15990
+    },
+    {
+      "epoch": 0.3768074984692195,
+      "grad_norm": 2.593400478363037,
+      "learning_rate": 0.00012467147096227216,
+      "loss": 2.1123,
+      "step": 16000
+    },
+    {
+      "epoch": 0.3770430031557628,
+      "grad_norm": 2.239337205886841,
+      "learning_rate": 0.0001246243700249635,
+      "loss": 2.2664,
+      "step": 16010
+    },
+    {
+      "epoch": 0.37727850784230604,
+      "grad_norm": 2.6145036220550537,
+      "learning_rate": 0.00012457726908765486,
+      "loss": 2.1313,
+      "step": 16020
+    },
+    {
+      "epoch": 0.3775140125288493,
+      "grad_norm": 2.3163740634918213,
+      "learning_rate": 0.0001245301681503462,
+      "loss": 2.1796,
+      "step": 16030
+    },
+    {
+      "epoch": 0.37774951721539257,
+      "grad_norm": 1.7631512880325317,
+      "learning_rate": 0.00012448306721303757,
+      "loss": 2.0957,
+      "step": 16040
+    },
+    {
+      "epoch": 0.37798502190193584,
+      "grad_norm": 1.9558098316192627,
+      "learning_rate": 0.0001244359662757289,
+      "loss": 2.0811,
+      "step": 16050
+    },
+    {
+      "epoch": 0.3782205265884791,
+      "grad_norm": 1.9874584674835205,
+      "learning_rate": 0.00012438886533842024,
+      "loss": 1.7742,
+      "step": 16060
+    },
+    {
+      "epoch": 0.37845603127502236,
+      "grad_norm": 2.0579781532287598,
+      "learning_rate": 0.0001243417644011116,
+      "loss": 1.9229,
+      "step": 16070
+    },
+    {
+      "epoch": 0.37869153596156563,
+      "grad_norm": 2.639272928237915,
+      "learning_rate": 0.00012429466346380294,
+      "loss": 2.294,
+      "step": 16080
+    },
+    {
+      "epoch": 0.3789270406481089,
+      "grad_norm": 2.7305397987365723,
+      "learning_rate": 0.0001242475625264943,
+      "loss": 2.1147,
+      "step": 16090
+    },
+    {
+      "epoch": 0.37916254533465216,
+      "grad_norm": 3.5461792945861816,
+      "learning_rate": 0.00012420046158918564,
+      "loss": 2.1957,
+      "step": 16100
+    },
+    {
+      "epoch": 0.3793980500211954,
+      "grad_norm": 2.2670328617095947,
+      "learning_rate": 0.00012415336065187696,
+      "loss": 2.0242,
+      "step": 16110
+    },
+    {
+      "epoch": 0.3796335547077387,
+      "grad_norm": 2.6801137924194336,
+      "learning_rate": 0.00012410625971456834,
+      "loss": 2.0179,
+      "step": 16120
+    },
+    {
+      "epoch": 0.37986905939428195,
+      "grad_norm": 2.0788025856018066,
+      "learning_rate": 0.00012405915877725966,
+      "loss": 1.8448,
+      "step": 16130
+    },
+    {
+      "epoch": 0.3801045640808252,
+      "grad_norm": 1.919297695159912,
+      "learning_rate": 0.00012401205783995101,
+      "loss": 1.9961,
+      "step": 16140
+    },
+    {
+      "epoch": 0.3803400687673685,
+      "grad_norm": 1.8758867979049683,
+      "learning_rate": 0.00012396495690264236,
+      "loss": 2.3458,
+      "step": 16150
+    },
+    {
+      "epoch": 0.38057557345391174,
+      "grad_norm": 2.559508800506592,
+      "learning_rate": 0.00012391785596533371,
+      "loss": 2.1396,
+      "step": 16160
+    },
+    {
+      "epoch": 0.380811078140455,
+      "grad_norm": 1.8591848611831665,
+      "learning_rate": 0.00012387075502802506,
+      "loss": 2.0795,
+      "step": 16170
+    },
+    {
+      "epoch": 0.38104658282699827,
+      "grad_norm": 1.8270467519760132,
+      "learning_rate": 0.00012382365409071641,
+      "loss": 1.9482,
+      "step": 16180
+    },
+    {
+      "epoch": 0.38128208751354153,
+      "grad_norm": 1.938462257385254,
+      "learning_rate": 0.00012377655315340774,
+      "loss": 2.0556,
+      "step": 16190
+    },
+    {
+      "epoch": 0.3815175922000848,
+      "grad_norm": 2.203063726425171,
+      "learning_rate": 0.00012372945221609912,
+      "loss": 2.1188,
+      "step": 16200
+    },
+    {
+      "epoch": 0.38175309688662806,
+      "grad_norm": 2.3647751808166504,
+      "learning_rate": 0.00012368235127879047,
+      "loss": 1.9052,
+      "step": 16210
+    },
+    {
+      "epoch": 0.3819886015731713,
+      "grad_norm": 2.129561185836792,
+      "learning_rate": 0.0001236352503414818,
+      "loss": 2.1535,
+      "step": 16220
+    },
+    {
+      "epoch": 0.3822241062597146,
+      "grad_norm": 2.4006335735321045,
+      "learning_rate": 0.00012358814940417317,
+      "loss": 2.0127,
+      "step": 16230
+    },
+    {
+      "epoch": 0.38245961094625786,
+      "grad_norm": 2.2458090782165527,
+      "learning_rate": 0.0001235410484668645,
+      "loss": 2.1334,
+      "step": 16240
+    },
+    {
+      "epoch": 0.3826951156328011,
+      "grad_norm": 2.183962106704712,
+      "learning_rate": 0.00012349394752955587,
+      "loss": 2.0084,
+      "step": 16250
+    },
+    {
+      "epoch": 0.38293062031934433,
+      "grad_norm": 2.067183017730713,
+      "learning_rate": 0.0001234468465922472,
+      "loss": 1.9644,
+      "step": 16260
+    },
+    {
+      "epoch": 0.3831661250058876,
+      "grad_norm": 1.8971290588378906,
+      "learning_rate": 0.00012339974565493854,
+      "loss": 1.9136,
+      "step": 16270
+    },
+    {
+      "epoch": 0.38340162969243086,
+      "grad_norm": 2.4257447719573975,
+      "learning_rate": 0.0001233526447176299,
+      "loss": 2.3739,
+      "step": 16280
+    },
+    {
+      "epoch": 0.3836371343789741,
+      "grad_norm": 2.2010610103607178,
+      "learning_rate": 0.00012330554378032124,
+      "loss": 2.1275,
+      "step": 16290
+    },
+    {
+      "epoch": 0.3838726390655174,
+      "grad_norm": 2.2034873962402344,
+      "learning_rate": 0.0001232584428430126,
+      "loss": 2.007,
+      "step": 16300
+    },
+    {
+      "epoch": 0.38410814375206065,
+      "grad_norm": 2.489978551864624,
+      "learning_rate": 0.00012321134190570394,
+      "loss": 2.0106,
+      "step": 16310
+    },
+    {
+      "epoch": 0.3843436484386039,
+      "grad_norm": 2.219968795776367,
+      "learning_rate": 0.00012316424096839526,
+      "loss": 1.946,
+      "step": 16320
+    },
+    {
+      "epoch": 0.3845791531251472,
+      "grad_norm": 3.080683946609497,
+      "learning_rate": 0.00012311714003108664,
+      "loss": 2.1341,
+      "step": 16330
+    },
+    {
+      "epoch": 0.38481465781169044,
+      "grad_norm": 2.717668294906616,
+      "learning_rate": 0.00012307003909377796,
+      "loss": 2.0214,
+      "step": 16340
+    },
+    {
+      "epoch": 0.3850501624982337,
+      "grad_norm": 2.038290023803711,
+      "learning_rate": 0.00012302293815646931,
+      "loss": 1.8628,
+      "step": 16350
+    },
+    {
+      "epoch": 0.38528566718477697,
+      "grad_norm": 2.27325177192688,
+      "learning_rate": 0.00012297583721916066,
+      "loss": 2.0714,
+      "step": 16360
+    },
+    {
+      "epoch": 0.38552117187132023,
+      "grad_norm": 2.770235776901245,
+      "learning_rate": 0.00012292873628185202,
+      "loss": 2.1704,
+      "step": 16370
+    },
+    {
+      "epoch": 0.3857566765578635,
+      "grad_norm": 2.967625379562378,
+      "learning_rate": 0.00012288163534454337,
+      "loss": 2.0807,
+      "step": 16380
+    },
+    {
+      "epoch": 0.38599218124440676,
+      "grad_norm": 2.048448085784912,
+      "learning_rate": 0.00012283453440723472,
+      "loss": 2.1044,
+      "step": 16390
+    },
+    {
+      "epoch": 0.38622768593095,
+      "grad_norm": 2.9608845710754395,
+      "learning_rate": 0.00012278743346992604,
+      "loss": 2.1001,
+      "step": 16400
+    },
+    {
+      "epoch": 0.3864631906174933,
+      "grad_norm": 1.9301419258117676,
+      "learning_rate": 0.00012274033253261742,
+      "loss": 2.0542,
+      "step": 16410
+    },
+    {
+      "epoch": 0.38669869530403655,
+      "grad_norm": 1.865768313407898,
+      "learning_rate": 0.00012269323159530874,
+      "loss": 2.001,
+      "step": 16420
+    },
+    {
+      "epoch": 0.3869341999905798,
+      "grad_norm": 2.2581074237823486,
+      "learning_rate": 0.0001226461306580001,
+      "loss": 2.0847,
+      "step": 16430
+    },
+    {
+      "epoch": 0.3871697046771231,
+      "grad_norm": 1.6205297708511353,
+      "learning_rate": 0.00012259902972069144,
+      "loss": 2.1412,
+      "step": 16440
+    },
+    {
+      "epoch": 0.38740520936366635,
+      "grad_norm": 2.4812774658203125,
+      "learning_rate": 0.0001225519287833828,
+      "loss": 2.1315,
+      "step": 16450
+    },
+    {
+      "epoch": 0.3876407140502096,
+      "grad_norm": 2.2032692432403564,
+      "learning_rate": 0.00012250482784607414,
+      "loss": 2.108,
+      "step": 16460
+    },
+    {
+      "epoch": 0.3878762187367529,
+      "grad_norm": 2.2076354026794434,
+      "learning_rate": 0.0001224577269087655,
+      "loss": 2.1367,
+      "step": 16470
+    },
+    {
+      "epoch": 0.38811172342329614,
+      "grad_norm": 2.061929225921631,
+      "learning_rate": 0.0001224106259714568,
+      "loss": 1.8931,
+      "step": 16480
+    },
+    {
+      "epoch": 0.3883472281098394,
+      "grad_norm": 1.9706366062164307,
+      "learning_rate": 0.0001223635250341482,
+      "loss": 2.1087,
+      "step": 16490
+    },
+    {
+      "epoch": 0.38858273279638267,
+      "grad_norm": 2.2798683643341064,
+      "learning_rate": 0.00012231642409683954,
+      "loss": 2.152,
+      "step": 16500
+    },
+    {
+      "epoch": 0.38881823748292593,
+      "grad_norm": 2.740971326828003,
+      "learning_rate": 0.0001222693231595309,
+      "loss": 1.9273,
+      "step": 16510
+    },
+    {
+      "epoch": 0.3890537421694692,
+      "grad_norm": 2.453092575073242,
+      "learning_rate": 0.00012222222222222224,
+      "loss": 1.8969,
+      "step": 16520
+    },
+    {
+      "epoch": 0.38928924685601246,
+      "grad_norm": 2.553710460662842,
+      "learning_rate": 0.00012217512128491356,
+      "loss": 2.0968,
+      "step": 16530
+    },
+    {
+      "epoch": 0.38952475154255567,
+      "grad_norm": 2.732374429702759,
+      "learning_rate": 0.00012212802034760494,
+      "loss": 2.2074,
+      "step": 16540
+    },
+    {
+      "epoch": 0.38976025622909893,
+      "grad_norm": 2.342780113220215,
+      "learning_rate": 0.00012208091941029627,
+      "loss": 2.0617,
+      "step": 16550
+    },
+    {
+      "epoch": 0.3899957609156422,
+      "grad_norm": 1.9254868030548096,
+      "learning_rate": 0.00012203381847298763,
+      "loss": 1.8196,
+      "step": 16560
+    },
+    {
+      "epoch": 0.39023126560218546,
+      "grad_norm": 1.8668605089187622,
+      "learning_rate": 0.00012198671753567897,
+      "loss": 2.1088,
+      "step": 16570
+    },
+    {
+      "epoch": 0.3904667702887287,
+      "grad_norm": 2.149210214614868,
+      "learning_rate": 0.00012193961659837032,
+      "loss": 2.232,
+      "step": 16580
+    },
+    {
+      "epoch": 0.390702274975272,
+      "grad_norm": 2.4574036598205566,
+      "learning_rate": 0.00012189251566106165,
+      "loss": 2.0236,
+      "step": 16590
+    },
+    {
+      "epoch": 0.39093777966181525,
+      "grad_norm": 1.9799808263778687,
+      "learning_rate": 0.00012184541472375302,
+      "loss": 2.1751,
+      "step": 16600
+    },
+    {
+      "epoch": 0.3911732843483585,
+      "grad_norm": 2.1461141109466553,
+      "learning_rate": 0.00012179831378644435,
+      "loss": 2.1239,
+      "step": 16610
+    },
+    {
+      "epoch": 0.3914087890349018,
+      "grad_norm": 2.3437371253967285,
+      "learning_rate": 0.0001217512128491357,
+      "loss": 2.2036,
+      "step": 16620
+    },
+    {
+      "epoch": 0.39164429372144505,
+      "grad_norm": 2.012018918991089,
+      "learning_rate": 0.00012170411191182704,
+      "loss": 2.1692,
+      "step": 16630
+    },
+    {
+      "epoch": 0.3918797984079883,
+      "grad_norm": 2.889059066772461,
+      "learning_rate": 0.0001216570109745184,
+      "loss": 2.2988,
+      "step": 16640
+    },
+    {
+      "epoch": 0.3921153030945316,
+      "grad_norm": 2.31487774848938,
+      "learning_rate": 0.00012160991003720974,
+      "loss": 2.1439,
+      "step": 16650
+    },
+    {
+      "epoch": 0.39235080778107484,
+      "grad_norm": 1.9940412044525146,
+      "learning_rate": 0.0001215628090999011,
+      "loss": 2.1955,
+      "step": 16660
+    },
+    {
+      "epoch": 0.3925863124676181,
+      "grad_norm": 2.186189889907837,
+      "learning_rate": 0.00012151570816259243,
+      "loss": 1.9478,
+      "step": 16670
+    },
+    {
+      "epoch": 0.39282181715416137,
+      "grad_norm": 1.8489048480987549,
+      "learning_rate": 0.00012146860722528379,
+      "loss": 2.0117,
+      "step": 16680
+    },
+    {
+      "epoch": 0.39305732184070463,
+      "grad_norm": 2.188730239868164,
+      "learning_rate": 0.00012142150628797513,
+      "loss": 2.367,
+      "step": 16690
+    },
+    {
+      "epoch": 0.3932928265272479,
+      "grad_norm": 2.6312897205352783,
+      "learning_rate": 0.00012137440535066649,
+      "loss": 2.1655,
+      "step": 16700
+    },
+    {
+      "epoch": 0.39352833121379116,
+      "grad_norm": 2.244814872741699,
+      "learning_rate": 0.00012132730441335783,
+      "loss": 2.0669,
+      "step": 16710
+    },
+    {
+      "epoch": 0.3937638359003344,
+      "grad_norm": 3.1631484031677246,
+      "learning_rate": 0.00012128020347604918,
+      "loss": 2.3648,
+      "step": 16720
+    },
+    {
+      "epoch": 0.3939993405868777,
+      "grad_norm": 2.074009418487549,
+      "learning_rate": 0.00012123310253874052,
+      "loss": 2.192,
+      "step": 16730
+    },
+    {
+      "epoch": 0.39423484527342095,
+      "grad_norm": 2.1103250980377197,
+      "learning_rate": 0.00012118600160143188,
+      "loss": 1.9476,
+      "step": 16740
+    },
+    {
+      "epoch": 0.3944703499599642,
+      "grad_norm": 2.7866339683532715,
+      "learning_rate": 0.00012113890066412322,
+      "loss": 1.9956,
+      "step": 16750
+    },
+    {
+      "epoch": 0.3947058546465075,
+      "grad_norm": 2.7085251808166504,
+      "learning_rate": 0.00012109179972681457,
+      "loss": 2.2952,
+      "step": 16760
+    },
+    {
+      "epoch": 0.39494135933305075,
+      "grad_norm": 1.9454951286315918,
+      "learning_rate": 0.0001210446987895059,
+      "loss": 2.1263,
+      "step": 16770
+    },
+    {
+      "epoch": 0.395176864019594,
+      "grad_norm": 3.369178056716919,
+      "learning_rate": 0.00012099759785219727,
+      "loss": 2.0465,
+      "step": 16780
+    },
+    {
+      "epoch": 0.3954123687061373,
+      "grad_norm": 2.1586029529571533,
+      "learning_rate": 0.00012095049691488862,
+      "loss": 2.1699,
+      "step": 16790
+    },
+    {
+      "epoch": 0.39564787339268054,
+      "grad_norm": 3.4209299087524414,
+      "learning_rate": 0.00012090339597757995,
+      "loss": 1.9791,
+      "step": 16800
+    },
+    {
+      "epoch": 0.3958833780792238,
+      "grad_norm": 2.1973490715026855,
+      "learning_rate": 0.00012085629504027132,
+      "loss": 2.1723,
+      "step": 16810
+    },
+    {
+      "epoch": 0.39611888276576707,
+      "grad_norm": 2.067542791366577,
+      "learning_rate": 0.00012080919410296265,
+      "loss": 1.8388,
+      "step": 16820
+    },
+    {
+      "epoch": 0.3963543874523103,
+      "grad_norm": 2.0267772674560547,
+      "learning_rate": 0.000120762093165654,
+      "loss": 2.1342,
+      "step": 16830
+    },
+    {
+      "epoch": 0.39658989213885354,
+      "grad_norm": 1.9653459787368774,
+      "learning_rate": 0.00012071499222834534,
+      "loss": 2.1676,
+      "step": 16840
+    },
+    {
+      "epoch": 0.3968253968253968,
+      "grad_norm": 2.003641366958618,
+      "learning_rate": 0.0001206678912910367,
+      "loss": 2.119,
+      "step": 16850
+    },
+    {
+      "epoch": 0.39706090151194007,
+      "grad_norm": 2.4826791286468506,
+      "learning_rate": 0.00012062079035372804,
+      "loss": 1.9719,
+      "step": 16860
+    },
+    {
+      "epoch": 0.39729640619848333,
+      "grad_norm": 2.9601070880889893,
+      "learning_rate": 0.0001205736894164194,
+      "loss": 1.9954,
+      "step": 16870
+    },
+    {
+      "epoch": 0.3975319108850266,
+      "grad_norm": 2.5066933631896973,
+      "learning_rate": 0.00012052658847911074,
+      "loss": 2.0153,
+      "step": 16880
+    },
+    {
+      "epoch": 0.39776741557156986,
+      "grad_norm": 2.712014675140381,
+      "learning_rate": 0.00012047948754180209,
+      "loss": 2.2096,
+      "step": 16890
+    },
+    {
+      "epoch": 0.3980029202581131,
+      "grad_norm": 1.7923743724822998,
+      "learning_rate": 0.00012043238660449343,
+      "loss": 2.2143,
+      "step": 16900
+    },
+    {
+      "epoch": 0.3982384249446564,
+      "grad_norm": 2.5782217979431152,
+      "learning_rate": 0.00012038528566718479,
+      "loss": 2.2435,
+      "step": 16910
+    },
+    {
+      "epoch": 0.39847392963119965,
+      "grad_norm": 2.224064350128174,
+      "learning_rate": 0.00012033818472987613,
+      "loss": 2.1383,
+      "step": 16920
+    },
+    {
+      "epoch": 0.3987094343177429,
+      "grad_norm": 2.3517062664031982,
+      "learning_rate": 0.00012029108379256748,
+      "loss": 1.7965,
+      "step": 16930
+    },
+    {
+      "epoch": 0.3989449390042862,
+      "grad_norm": 3.8429908752441406,
+      "learning_rate": 0.00012024398285525882,
+      "loss": 2.0901,
+      "step": 16940
+    },
+    {
+      "epoch": 0.39918044369082945,
+      "grad_norm": 2.49701189994812,
+      "learning_rate": 0.00012019688191795018,
+      "loss": 2.075,
+      "step": 16950
+    },
+    {
+      "epoch": 0.3994159483773727,
+      "grad_norm": 2.411953926086426,
+      "learning_rate": 0.00012014978098064152,
+      "loss": 2.2769,
+      "step": 16960
+    },
+    {
+      "epoch": 0.399651453063916,
+      "grad_norm": 2.9988455772399902,
+      "learning_rate": 0.00012010268004333287,
+      "loss": 1.979,
+      "step": 16970
+    },
+    {
+      "epoch": 0.39988695775045924,
+      "grad_norm": 3.0603721141815186,
+      "learning_rate": 0.0001200555791060242,
+      "loss": 2.0761,
+      "step": 16980
+    },
+    {
+      "epoch": 0.4001224624370025,
+      "grad_norm": 2.097357988357544,
+      "learning_rate": 0.00012000847816871557,
+      "loss": 2.1266,
+      "step": 16990
+    },
+    {
+      "epoch": 0.40035796712354577,
+      "grad_norm": 2.6588997840881348,
+      "learning_rate": 0.0001199613772314069,
+      "loss": 2.1037,
+      "step": 17000
+    },
+    {
+      "epoch": 0.40059347181008903,
+      "grad_norm": 2.0685205459594727,
+      "learning_rate": 0.00011991427629409825,
+      "loss": 2.0696,
+      "step": 17010
+    },
+    {
+      "epoch": 0.4008289764966323,
+      "grad_norm": 2.6148645877838135,
+      "learning_rate": 0.00011986717535678959,
+      "loss": 2.0947,
+      "step": 17020
+    },
+    {
+      "epoch": 0.40106448118317556,
+      "grad_norm": 2.0467002391815186,
+      "learning_rate": 0.00011982007441948096,
+      "loss": 2.2171,
+      "step": 17030
+    },
+    {
+      "epoch": 0.4012999858697188,
+      "grad_norm": 2.4288554191589355,
+      "learning_rate": 0.00011977297348217229,
+      "loss": 2.1036,
+      "step": 17040
+    },
+    {
+      "epoch": 0.4015354905562621,
+      "grad_norm": 2.3127670288085938,
+      "learning_rate": 0.00011972587254486364,
+      "loss": 2.0517,
+      "step": 17050
+    },
+    {
+      "epoch": 0.40177099524280535,
+      "grad_norm": 2.3823368549346924,
+      "learning_rate": 0.000119678771607555,
+      "loss": 2.0046,
+      "step": 17060
+    },
+    {
+      "epoch": 0.4020064999293486,
+      "grad_norm": 2.0719432830810547,
+      "learning_rate": 0.00011963167067024634,
+      "loss": 2.0568,
+      "step": 17070
+    },
+    {
+      "epoch": 0.4022420046158919,
+      "grad_norm": 2.4580512046813965,
+      "learning_rate": 0.0001195845697329377,
+      "loss": 2.0559,
+      "step": 17080
+    },
+    {
+      "epoch": 0.40247750930243514,
+      "grad_norm": 2.124021291732788,
+      "learning_rate": 0.00011953746879562904,
+      "loss": 2.086,
+      "step": 17090
+    },
+    {
+      "epoch": 0.4027130139889784,
+      "grad_norm": 2.4031240940093994,
+      "learning_rate": 0.00011949036785832039,
+      "loss": 2.1624,
+      "step": 17100
+    },
+    {
+      "epoch": 0.4029485186755216,
+      "grad_norm": 2.5331597328186035,
+      "learning_rate": 0.00011944326692101173,
+      "loss": 2.0293,
+      "step": 17110
+    },
+    {
+      "epoch": 0.4031840233620649,
+      "grad_norm": 2.3105568885803223,
+      "learning_rate": 0.0001193961659837031,
+      "loss": 2.0536,
+      "step": 17120
+    },
+    {
+      "epoch": 0.40341952804860814,
+      "grad_norm": 1.935504674911499,
+      "learning_rate": 0.00011934906504639443,
+      "loss": 2.0363,
+      "step": 17130
+    },
+    {
+      "epoch": 0.4036550327351514,
+      "grad_norm": 1.818901777267456,
+      "learning_rate": 0.00011930196410908578,
+      "loss": 2.1138,
+      "step": 17140
+    },
+    {
+      "epoch": 0.4038905374216947,
+      "grad_norm": 2.2937822341918945,
+      "learning_rate": 0.00011925486317177712,
+      "loss": 2.2075,
+      "step": 17150
+    },
+    {
+      "epoch": 0.40412604210823794,
+      "grad_norm": 2.4323720932006836,
+      "learning_rate": 0.00011920776223446848,
+      "loss": 1.8704,
+      "step": 17160
+    },
+    {
+      "epoch": 0.4043615467947812,
+      "grad_norm": 2.0773727893829346,
+      "learning_rate": 0.00011916066129715982,
+      "loss": 2.1124,
+      "step": 17170
+    },
+    {
+      "epoch": 0.40459705148132447,
+      "grad_norm": 2.424532413482666,
+      "learning_rate": 0.00011911356035985117,
+      "loss": 1.9422,
+      "step": 17180
+    },
+    {
+      "epoch": 0.40483255616786773,
+      "grad_norm": 2.1791512966156006,
+      "learning_rate": 0.0001190664594225425,
+      "loss": 2.0547,
+      "step": 17190
+    },
+    {
+      "epoch": 0.405068060854411,
+      "grad_norm": 2.939730167388916,
+      "learning_rate": 0.00011901935848523387,
+      "loss": 1.9766,
+      "step": 17200
+    },
+    {
+      "epoch": 0.40530356554095426,
+      "grad_norm": 2.1646764278411865,
+      "learning_rate": 0.0001189722575479252,
+      "loss": 1.9632,
+      "step": 17210
+    },
+    {
+      "epoch": 0.4055390702274975,
+      "grad_norm": 3.135854721069336,
+      "learning_rate": 0.00011892515661061656,
+      "loss": 1.942,
+      "step": 17220
+    },
+    {
+      "epoch": 0.4057745749140408,
+      "grad_norm": 2.3029401302337646,
+      "learning_rate": 0.00011887805567330789,
+      "loss": 2.3985,
+      "step": 17230
+    },
+    {
+      "epoch": 0.40601007960058405,
+      "grad_norm": 2.5653719902038574,
+      "learning_rate": 0.00011883095473599926,
+      "loss": 2.125,
+      "step": 17240
+    },
+    {
+      "epoch": 0.4062455842871273,
+      "grad_norm": 2.4413366317749023,
+      "learning_rate": 0.00011878385379869059,
+      "loss": 2.2774,
+      "step": 17250
+    },
+    {
+      "epoch": 0.4064810889736706,
+      "grad_norm": 1.920428991317749,
+      "learning_rate": 0.00011873675286138196,
+      "loss": 2.3723,
+      "step": 17260
+    },
+    {
+      "epoch": 0.40671659366021384,
+      "grad_norm": 2.025369644165039,
+      "learning_rate": 0.00011868965192407328,
+      "loss": 2.1473,
+      "step": 17270
+    },
+    {
+      "epoch": 0.4069520983467571,
+      "grad_norm": 2.081719160079956,
+      "learning_rate": 0.00011864255098676464,
+      "loss": 2.2024,
+      "step": 17280
+    },
+    {
+      "epoch": 0.40718760303330037,
+      "grad_norm": 1.9507285356521606,
+      "learning_rate": 0.00011859545004945598,
+      "loss": 2.0511,
+      "step": 17290
+    },
+    {
+      "epoch": 0.40742310771984364,
+      "grad_norm": 2.085679531097412,
+      "learning_rate": 0.00011854834911214734,
+      "loss": 2.0069,
+      "step": 17300
+    },
+    {
+      "epoch": 0.4076586124063869,
+      "grad_norm": 2.4207077026367188,
+      "learning_rate": 0.00011850124817483868,
+      "loss": 2.1838,
+      "step": 17310
+    },
+    {
+      "epoch": 0.40789411709293016,
+      "grad_norm": 1.945299744606018,
+      "learning_rate": 0.00011845414723753003,
+      "loss": 2.1365,
+      "step": 17320
+    },
+    {
+      "epoch": 0.40812962177947343,
+      "grad_norm": 2.0724103450775146,
+      "learning_rate": 0.00011840704630022137,
+      "loss": 2.1791,
+      "step": 17330
+    },
+    {
+      "epoch": 0.4083651264660167,
+      "grad_norm": 2.00799298286438,
+      "learning_rate": 0.00011835994536291273,
+      "loss": 2.103,
+      "step": 17340
+    },
+    {
+      "epoch": 0.40860063115255996,
+      "grad_norm": 2.6130778789520264,
+      "learning_rate": 0.00011831284442560408,
+      "loss": 2.1468,
+      "step": 17350
+    },
+    {
+      "epoch": 0.4088361358391032,
+      "grad_norm": 2.042046308517456,
+      "learning_rate": 0.00011826574348829542,
+      "loss": 1.9138,
+      "step": 17360
+    },
+    {
+      "epoch": 0.4090716405256465,
+      "grad_norm": 2.197761058807373,
+      "learning_rate": 0.00011821864255098678,
+      "loss": 2.0824,
+      "step": 17370
+    },
+    {
+      "epoch": 0.40930714521218975,
+      "grad_norm": 2.178068161010742,
+      "learning_rate": 0.00011817154161367812,
+      "loss": 2.0455,
+      "step": 17380
+    },
+    {
+      "epoch": 0.40954264989873296,
+      "grad_norm": 1.9603676795959473,
+      "learning_rate": 0.00011812444067636947,
+      "loss": 2.0522,
+      "step": 17390
+    },
+    {
+      "epoch": 0.4097781545852762,
+      "grad_norm": 1.8290703296661377,
+      "learning_rate": 0.0001180773397390608,
+      "loss": 2.0019,
+      "step": 17400
+    },
+    {
+      "epoch": 0.4100136592718195,
+      "grad_norm": 2.015998601913452,
+      "learning_rate": 0.00011803023880175217,
+      "loss": 2.2494,
+      "step": 17410
+    },
+    {
+      "epoch": 0.41024916395836275,
+      "grad_norm": 2.149958610534668,
+      "learning_rate": 0.0001179831378644435,
+      "loss": 1.9823,
+      "step": 17420
+    },
+    {
+      "epoch": 0.410484668644906,
+      "grad_norm": 1.841565489768982,
+      "learning_rate": 0.00011793603692713486,
+      "loss": 1.9568,
+      "step": 17430
+    },
+    {
+      "epoch": 0.4107201733314493,
+      "grad_norm": 2.855804204940796,
+      "learning_rate": 0.00011788893598982619,
+      "loss": 2.1627,
+      "step": 17440
+    },
+    {
+      "epoch": 0.41095567801799254,
+      "grad_norm": 2.0609381198883057,
+      "learning_rate": 0.00011784183505251756,
+      "loss": 2.0022,
+      "step": 17450
+    },
+    {
+      "epoch": 0.4111911827045358,
+      "grad_norm": 2.362701416015625,
+      "learning_rate": 0.0001177947341152089,
+      "loss": 2.2441,
+      "step": 17460
+    },
+    {
+      "epoch": 0.41142668739107907,
+      "grad_norm": 1.9817253351211548,
+      "learning_rate": 0.00011774763317790026,
+      "loss": 1.8491,
+      "step": 17470
+    },
+    {
+      "epoch": 0.41166219207762234,
+      "grad_norm": 1.9875420331954956,
+      "learning_rate": 0.0001177005322405916,
+      "loss": 1.8271,
+      "step": 17480
+    },
+    {
+      "epoch": 0.4118976967641656,
+      "grad_norm": 1.8128178119659424,
+      "learning_rate": 0.00011765343130328294,
+      "loss": 2.0102,
+      "step": 17490
+    },
+    {
+      "epoch": 0.41213320145070886,
+      "grad_norm": 2.7656726837158203,
+      "learning_rate": 0.00011760633036597428,
+      "loss": 2.0932,
+      "step": 17500
+    },
+    {
+      "epoch": 0.41236870613725213,
+      "grad_norm": 2.31899094581604,
+      "learning_rate": 0.00011755922942866564,
+      "loss": 2.3904,
+      "step": 17510
+    },
+    {
+      "epoch": 0.4126042108237954,
+      "grad_norm": 1.7432920932769775,
+      "learning_rate": 0.00011751212849135698,
+      "loss": 2.0688,
+      "step": 17520
+    },
+    {
+      "epoch": 0.41283971551033866,
+      "grad_norm": 2.6849148273468018,
+      "learning_rate": 0.00011746502755404833,
+      "loss": 1.9904,
+      "step": 17530
+    },
+    {
+      "epoch": 0.4130752201968819,
+      "grad_norm": 2.1877646446228027,
+      "learning_rate": 0.00011741792661673967,
+      "loss": 2.1276,
+      "step": 17540
+    },
+    {
+      "epoch": 0.4133107248834252,
+      "grad_norm": 2.3878908157348633,
+      "learning_rate": 0.00011737082567943103,
+      "loss": 2.0121,
+      "step": 17550
+    },
+    {
+      "epoch": 0.41354622956996845,
+      "grad_norm": 2.5052924156188965,
+      "learning_rate": 0.00011732372474212237,
+      "loss": 2.1646,
+      "step": 17560
+    },
+    {
+      "epoch": 0.4137817342565117,
+      "grad_norm": 1.9085092544555664,
+      "learning_rate": 0.00011727662380481372,
+      "loss": 2.0815,
+      "step": 17570
+    },
+    {
+      "epoch": 0.414017238943055,
+      "grad_norm": 2.4492745399475098,
+      "learning_rate": 0.00011722952286750506,
+      "loss": 2.0763,
+      "step": 17580
+    },
+    {
+      "epoch": 0.41425274362959824,
+      "grad_norm": 2.189110279083252,
+      "learning_rate": 0.00011718242193019642,
+      "loss": 2.114,
+      "step": 17590
+    },
+    {
+      "epoch": 0.4144882483161415,
+      "grad_norm": 2.4104158878326416,
+      "learning_rate": 0.00011713532099288776,
+      "loss": 2.19,
+      "step": 17600
+    },
+    {
+      "epoch": 0.41472375300268477,
+      "grad_norm": 1.884446620941162,
+      "learning_rate": 0.0001170882200555791,
+      "loss": 1.9784,
+      "step": 17610
+    },
+    {
+      "epoch": 0.41495925768922803,
+      "grad_norm": 2.4759879112243652,
+      "learning_rate": 0.00011704111911827047,
+      "loss": 2.2371,
+      "step": 17620
+    },
+    {
+      "epoch": 0.4151947623757713,
+      "grad_norm": 2.4873902797698975,
+      "learning_rate": 0.00011699401818096181,
+      "loss": 2.1282,
+      "step": 17630
+    },
+    {
+      "epoch": 0.41543026706231456,
+      "grad_norm": 1.9872322082519531,
+      "learning_rate": 0.00011694691724365317,
+      "loss": 2.1676,
+      "step": 17640
+    },
+    {
+      "epoch": 0.4156657717488578,
+      "grad_norm": 2.117286205291748,
+      "learning_rate": 0.0001168998163063445,
+      "loss": 2.1725,
+      "step": 17650
+    },
+    {
+      "epoch": 0.4159012764354011,
+      "grad_norm": 2.261207342147827,
+      "learning_rate": 0.00011685271536903586,
+      "loss": 2.1271,
+      "step": 17660
+    },
+    {
+      "epoch": 0.4161367811219443,
+      "grad_norm": 2.076218843460083,
+      "learning_rate": 0.0001168056144317272,
+      "loss": 1.926,
+      "step": 17670
+    },
+    {
+      "epoch": 0.41637228580848756,
+      "grad_norm": 2.5241241455078125,
+      "learning_rate": 0.00011675851349441856,
+      "loss": 1.9504,
+      "step": 17680
+    },
+    {
+      "epoch": 0.4166077904950308,
+      "grad_norm": 2.2576680183410645,
+      "learning_rate": 0.0001167114125571099,
+      "loss": 2.2149,
+      "step": 17690
+    },
+    {
+      "epoch": 0.4168432951815741,
+      "grad_norm": 2.1343159675598145,
+      "learning_rate": 0.00011666431161980125,
+      "loss": 2.1592,
+      "step": 17700
+    },
+    {
+      "epoch": 0.41707879986811736,
+      "grad_norm": 2.8269872665405273,
+      "learning_rate": 0.00011661721068249258,
+      "loss": 1.8489,
+      "step": 17710
+    },
+    {
+      "epoch": 0.4173143045546606,
+      "grad_norm": 2.6557393074035645,
+      "learning_rate": 0.00011657010974518395,
+      "loss": 2.3097,
+      "step": 17720
+    },
+    {
+      "epoch": 0.4175498092412039,
+      "grad_norm": 1.9406561851501465,
+      "learning_rate": 0.00011652300880787528,
+      "loss": 2.2424,
+      "step": 17730
+    },
+    {
+      "epoch": 0.41778531392774715,
+      "grad_norm": 2.8886189460754395,
+      "learning_rate": 0.00011647590787056663,
+      "loss": 1.9974,
+      "step": 17740
+    },
+    {
+      "epoch": 0.4180208186142904,
+      "grad_norm": 2.958347797393799,
+      "learning_rate": 0.00011642880693325797,
+      "loss": 2.0821,
+      "step": 17750
+    },
+    {
+      "epoch": 0.4182563233008337,
+      "grad_norm": 2.3907742500305176,
+      "learning_rate": 0.00011638170599594933,
+      "loss": 2.2051,
+      "step": 17760
+    },
+    {
+      "epoch": 0.41849182798737694,
+      "grad_norm": 2.605231523513794,
+      "learning_rate": 0.00011633460505864067,
+      "loss": 2.1911,
+      "step": 17770
+    },
+    {
+      "epoch": 0.4187273326739202,
+      "grad_norm": 2.1181795597076416,
+      "learning_rate": 0.00011628750412133202,
+      "loss": 1.8307,
+      "step": 17780
+    },
+    {
+      "epoch": 0.41896283736046347,
+      "grad_norm": 2.3362369537353516,
+      "learning_rate": 0.00011624040318402336,
+      "loss": 1.9971,
+      "step": 17790
+    },
+    {
+      "epoch": 0.41919834204700673,
+      "grad_norm": 2.4740500450134277,
+      "learning_rate": 0.00011619330224671472,
+      "loss": 2.0515,
+      "step": 17800
+    },
+    {
+      "epoch": 0.41943384673355,
+      "grad_norm": 2.383115768432617,
+      "learning_rate": 0.00011614620130940606,
+      "loss": 2.1157,
+      "step": 17810
+    },
+    {
+      "epoch": 0.41966935142009326,
+      "grad_norm": 1.487096905708313,
+      "learning_rate": 0.00011609910037209741,
+      "loss": 2.0911,
+      "step": 17820
+    },
+    {
+      "epoch": 0.4199048561066365,
+      "grad_norm": 2.061518430709839,
+      "learning_rate": 0.00011605199943478874,
+      "loss": 2.2484,
+      "step": 17830
+    },
+    {
+      "epoch": 0.4201403607931798,
+      "grad_norm": 2.2005579471588135,
+      "learning_rate": 0.00011600489849748011,
+      "loss": 2.0508,
+      "step": 17840
+    },
+    {
+      "epoch": 0.42037586547972305,
+      "grad_norm": 2.021574020385742,
+      "learning_rate": 0.00011595779756017144,
+      "loss": 2.0133,
+      "step": 17850
+    },
+    {
+      "epoch": 0.4206113701662663,
+      "grad_norm": 2.3463683128356934,
+      "learning_rate": 0.00011591069662286281,
+      "loss": 1.9324,
+      "step": 17860
+    },
+    {
+      "epoch": 0.4208468748528096,
+      "grad_norm": 2.0511763095855713,
+      "learning_rate": 0.00011586359568555413,
+      "loss": 1.8826,
+      "step": 17870
+    },
+    {
+      "epoch": 0.42108237953935285,
+      "grad_norm": 1.8911633491516113,
+      "learning_rate": 0.0001158164947482455,
+      "loss": 1.9932,
+      "step": 17880
+    },
+    {
+      "epoch": 0.4213178842258961,
+      "grad_norm": 2.640658140182495,
+      "learning_rate": 0.00011576939381093683,
+      "loss": 2.0184,
+      "step": 17890
+    },
+    {
+      "epoch": 0.4215533889124394,
+      "grad_norm": 2.112999677658081,
+      "learning_rate": 0.0001157222928736282,
+      "loss": 2.2726,
+      "step": 17900
+    },
+    {
+      "epoch": 0.42178889359898264,
+      "grad_norm": 2.526036262512207,
+      "learning_rate": 0.00011567519193631955,
+      "loss": 2.1118,
+      "step": 17910
+    },
+    {
+      "epoch": 0.4220243982855259,
+      "grad_norm": 2.948558807373047,
+      "learning_rate": 0.00011562809099901088,
+      "loss": 2.2965,
+      "step": 17920
+    },
+    {
+      "epoch": 0.42225990297206917,
+      "grad_norm": 2.221510648727417,
+      "learning_rate": 0.00011558099006170225,
+      "loss": 2.0943,
+      "step": 17930
+    },
+    {
+      "epoch": 0.42249540765861243,
+      "grad_norm": 1.9982761144638062,
+      "learning_rate": 0.00011553388912439358,
+      "loss": 2.1344,
+      "step": 17940
+    },
+    {
+      "epoch": 0.4227309123451557,
+      "grad_norm": 1.816463589668274,
+      "learning_rate": 0.00011548678818708493,
+      "loss": 2.1454,
+      "step": 17950
+    },
+    {
+      "epoch": 0.4229664170316989,
+      "grad_norm": 1.9133336544036865,
+      "learning_rate": 0.00011543968724977627,
+      "loss": 2.2219,
+      "step": 17960
+    },
+    {
+      "epoch": 0.42320192171824217,
+      "grad_norm": 2.87747859954834,
+      "learning_rate": 0.00011539258631246763,
+      "loss": 2.2701,
+      "step": 17970
+    },
+    {
+      "epoch": 0.42343742640478543,
+      "grad_norm": 2.2570955753326416,
+      "learning_rate": 0.00011534548537515897,
+      "loss": 2.2363,
+      "step": 17980
+    },
+    {
+      "epoch": 0.4236729310913287,
+      "grad_norm": 2.1838762760162354,
+      "learning_rate": 0.00011529838443785032,
+      "loss": 2.1544,
+      "step": 17990
+    },
+    {
+      "epoch": 0.42390843577787196,
+      "grad_norm": 2.118103504180908,
+      "learning_rate": 0.00011525128350054166,
+      "loss": 2.2437,
+      "step": 18000
+    },
+    {
+      "epoch": 0.4241439404644152,
+      "grad_norm": 1.956613540649414,
+      "learning_rate": 0.00011520418256323302,
+      "loss": 2.0544,
+      "step": 18010
+    },
+    {
+      "epoch": 0.4243794451509585,
+      "grad_norm": 2.5147013664245605,
+      "learning_rate": 0.00011515708162592436,
+      "loss": 2.1949,
+      "step": 18020
+    },
+    {
+      "epoch": 0.42461494983750175,
+      "grad_norm": 1.6883703470230103,
+      "learning_rate": 0.00011510998068861571,
+      "loss": 2.1606,
+      "step": 18030
+    },
+    {
+      "epoch": 0.424850454524045,
+      "grad_norm": 3.019367218017578,
+      "learning_rate": 0.00011506287975130705,
+      "loss": 2.1466,
+      "step": 18040
+    },
+    {
+      "epoch": 0.4250859592105883,
+      "grad_norm": 2.5443851947784424,
+      "learning_rate": 0.00011501577881399841,
+      "loss": 2.0999,
+      "step": 18050
+    },
+    {
+      "epoch": 0.42532146389713155,
+      "grad_norm": 1.759558081626892,
+      "learning_rate": 0.00011496867787668975,
+      "loss": 2.0157,
+      "step": 18060
+    },
+    {
+      "epoch": 0.4255569685836748,
+      "grad_norm": 2.4723591804504395,
+      "learning_rate": 0.00011492157693938111,
+      "loss": 2.2185,
+      "step": 18070
+    },
+    {
+      "epoch": 0.4257924732702181,
+      "grad_norm": 2.276010274887085,
+      "learning_rate": 0.00011487447600207245,
+      "loss": 2.197,
+      "step": 18080
+    },
+    {
+      "epoch": 0.42602797795676134,
+      "grad_norm": 2.547868013381958,
+      "learning_rate": 0.0001148273750647638,
+      "loss": 2.0393,
+      "step": 18090
+    },
+    {
+      "epoch": 0.4262634826433046,
+      "grad_norm": 2.285099983215332,
+      "learning_rate": 0.00011478027412745513,
+      "loss": 1.9714,
+      "step": 18100
+    },
+    {
+      "epoch": 0.42649898732984787,
+      "grad_norm": 1.7757441997528076,
+      "learning_rate": 0.0001147331731901465,
+      "loss": 2.0319,
+      "step": 18110
+    },
+    {
+      "epoch": 0.42673449201639113,
+      "grad_norm": 2.2466018199920654,
+      "learning_rate": 0.00011468607225283783,
+      "loss": 2.0001,
+      "step": 18120
+    },
+    {
+      "epoch": 0.4269699967029344,
+      "grad_norm": 2.274280548095703,
+      "learning_rate": 0.00011463897131552918,
+      "loss": 2.0503,
+      "step": 18130
+    },
+    {
+      "epoch": 0.42720550138947766,
+      "grad_norm": 2.588524341583252,
+      "learning_rate": 0.00011459187037822052,
+      "loss": 2.3182,
+      "step": 18140
+    },
+    {
+      "epoch": 0.4274410060760209,
+      "grad_norm": 2.081986904144287,
+      "learning_rate": 0.00011454476944091188,
+      "loss": 2.1953,
+      "step": 18150
+    },
+    {
+      "epoch": 0.4276765107625642,
+      "grad_norm": 2.046691417694092,
+      "learning_rate": 0.00011449766850360322,
+      "loss": 2.1025,
+      "step": 18160
+    },
+    {
+      "epoch": 0.42791201544910745,
+      "grad_norm": 2.94069504737854,
+      "learning_rate": 0.00011445056756629457,
+      "loss": 2.2368,
+      "step": 18170
+    },
+    {
+      "epoch": 0.4281475201356507,
+      "grad_norm": 2.245175361633301,
+      "learning_rate": 0.00011440346662898591,
+      "loss": 2.1502,
+      "step": 18180
+    },
+    {
+      "epoch": 0.428383024822194,
+      "grad_norm": 2.029442310333252,
+      "learning_rate": 0.00011435636569167727,
+      "loss": 2.2961,
+      "step": 18190
+    },
+    {
+      "epoch": 0.42861852950873724,
+      "grad_norm": 2.311154842376709,
+      "learning_rate": 0.00011430926475436862,
+      "loss": 2.0729,
+      "step": 18200
+    },
+    {
+      "epoch": 0.4288540341952805,
+      "grad_norm": 2.393167018890381,
+      "learning_rate": 0.00011426216381705996,
+      "loss": 2.0608,
+      "step": 18210
+    },
+    {
+      "epoch": 0.4290895388818238,
+      "grad_norm": 2.606497049331665,
+      "learning_rate": 0.00011421506287975132,
+      "loss": 2.0255,
+      "step": 18220
+    },
+    {
+      "epoch": 0.42932504356836704,
+      "grad_norm": 1.9125550985336304,
+      "learning_rate": 0.00011416796194244266,
+      "loss": 2.0945,
+      "step": 18230
+    },
+    {
+      "epoch": 0.42956054825491025,
+      "grad_norm": 2.4020402431488037,
+      "learning_rate": 0.00011412086100513402,
+      "loss": 2.0022,
+      "step": 18240
+    },
+    {
+      "epoch": 0.4297960529414535,
+      "grad_norm": 1.5923528671264648,
+      "learning_rate": 0.00011407376006782535,
+      "loss": 2.0348,
+      "step": 18250
+    },
+    {
+      "epoch": 0.4300315576279968,
+      "grad_norm": 2.0022480487823486,
+      "learning_rate": 0.00011402665913051671,
+      "loss": 2.1634,
+      "step": 18260
+    },
+    {
+      "epoch": 0.43026706231454004,
+      "grad_norm": 2.2088112831115723,
+      "learning_rate": 0.00011397955819320805,
+      "loss": 2.2046,
+      "step": 18270
+    },
+    {
+      "epoch": 0.4305025670010833,
+      "grad_norm": 1.8067007064819336,
+      "learning_rate": 0.00011393245725589941,
+      "loss": 2.0242,
+      "step": 18280
+    },
+    {
+      "epoch": 0.43073807168762657,
+      "grad_norm": 2.725207567214966,
+      "learning_rate": 0.00011388535631859075,
+      "loss": 1.8446,
+      "step": 18290
+    },
+    {
+      "epoch": 0.43097357637416983,
+      "grad_norm": 3.046445608139038,
+      "learning_rate": 0.0001138382553812821,
+      "loss": 2.1722,
+      "step": 18300
+    },
+    {
+      "epoch": 0.4312090810607131,
+      "grad_norm": 2.3470003604888916,
+      "learning_rate": 0.00011379115444397343,
+      "loss": 2.0065,
+      "step": 18310
+    },
+    {
+      "epoch": 0.43144458574725636,
+      "grad_norm": 3.09189772605896,
+      "learning_rate": 0.0001137440535066648,
+      "loss": 1.9948,
+      "step": 18320
+    },
+    {
+      "epoch": 0.4316800904337996,
+      "grad_norm": 2.13586163520813,
+      "learning_rate": 0.00011369695256935613,
+      "loss": 1.9383,
+      "step": 18330
+    },
+    {
+      "epoch": 0.4319155951203429,
+      "grad_norm": 2.6252524852752686,
+      "learning_rate": 0.00011364985163204748,
+      "loss": 1.9324,
+      "step": 18340
+    },
+    {
+      "epoch": 0.43215109980688615,
+      "grad_norm": 2.855048179626465,
+      "learning_rate": 0.00011360275069473882,
+      "loss": 2.3409,
+      "step": 18350
+    },
+    {
+      "epoch": 0.4323866044934294,
+      "grad_norm": 2.435730457305908,
+      "learning_rate": 0.00011355564975743019,
+      "loss": 2.1404,
+      "step": 18360
+    },
+    {
+      "epoch": 0.4326221091799727,
+      "grad_norm": 2.411512613296509,
+      "learning_rate": 0.00011350854882012152,
+      "loss": 2.2621,
+      "step": 18370
+    },
+    {
+      "epoch": 0.43285761386651594,
+      "grad_norm": 2.974522590637207,
+      "learning_rate": 0.00011346144788281287,
+      "loss": 2.1951,
+      "step": 18380
+    },
+    {
+      "epoch": 0.4330931185530592,
+      "grad_norm": 1.7726364135742188,
+      "learning_rate": 0.00011341434694550421,
+      "loss": 1.8892,
+      "step": 18390
+    },
+    {
+      "epoch": 0.4333286232396025,
+      "grad_norm": 2.825831174850464,
+      "learning_rate": 0.00011336724600819557,
+      "loss": 2.1635,
+      "step": 18400
+    },
+    {
+      "epoch": 0.43356412792614574,
+      "grad_norm": 2.5960309505462646,
+      "learning_rate": 0.00011332014507088691,
+      "loss": 2.0728,
+      "step": 18410
+    },
+    {
+      "epoch": 0.433799632612689,
+      "grad_norm": 2.0427327156066895,
+      "learning_rate": 0.00011327304413357826,
+      "loss": 2.0275,
+      "step": 18420
+    },
+    {
+      "epoch": 0.43403513729923227,
+      "grad_norm": 2.6198508739471436,
+      "learning_rate": 0.0001132259431962696,
+      "loss": 2.0241,
+      "step": 18430
+    },
+    {
+      "epoch": 0.43427064198577553,
+      "grad_norm": 2.224815845489502,
+      "learning_rate": 0.00011317884225896096,
+      "loss": 2.0585,
+      "step": 18440
+    },
+    {
+      "epoch": 0.4345061466723188,
+      "grad_norm": 2.4373631477355957,
+      "learning_rate": 0.0001131317413216523,
+      "loss": 2.0333,
+      "step": 18450
+    },
+    {
+      "epoch": 0.43474165135886206,
+      "grad_norm": 1.711761713027954,
+      "learning_rate": 0.00011308464038434366,
+      "loss": 1.7788,
+      "step": 18460
+    },
+    {
+      "epoch": 0.4349771560454053,
+      "grad_norm": 3.0095267295837402,
+      "learning_rate": 0.00011303753944703501,
+      "loss": 2.2233,
+      "step": 18470
+    },
+    {
+      "epoch": 0.4352126607319486,
+      "grad_norm": 2.0507631301879883,
+      "learning_rate": 0.00011299043850972635,
+      "loss": 2.019,
+      "step": 18480
+    },
+    {
+      "epoch": 0.43544816541849185,
+      "grad_norm": 2.733670711517334,
+      "learning_rate": 0.00011294333757241771,
+      "loss": 2.074,
+      "step": 18490
+    },
+    {
+      "epoch": 0.4356836701050351,
+      "grad_norm": 2.1250648498535156,
+      "learning_rate": 0.00011289623663510905,
+      "loss": 2.0513,
+      "step": 18500
+    },
+    {
+      "epoch": 0.4359191747915784,
+      "grad_norm": 3.0923218727111816,
+      "learning_rate": 0.0001128491356978004,
+      "loss": 2.3304,
+      "step": 18510
+    },
+    {
+      "epoch": 0.4361546794781216,
+      "grad_norm": 3.0069260597229004,
+      "learning_rate": 0.00011280203476049173,
+      "loss": 2.2824,
+      "step": 18520
+    },
+    {
+      "epoch": 0.43639018416466485,
+      "grad_norm": 2.1305959224700928,
+      "learning_rate": 0.0001127549338231831,
+      "loss": 2.0499,
+      "step": 18530
+    },
+    {
+      "epoch": 0.4366256888512081,
+      "grad_norm": 2.1745736598968506,
+      "learning_rate": 0.00011270783288587444,
+      "loss": 1.9903,
+      "step": 18540
+    },
+    {
+      "epoch": 0.4368611935377514,
+      "grad_norm": 2.113219738006592,
+      "learning_rate": 0.00011266073194856579,
+      "loss": 2.1466,
+      "step": 18550
+    },
+    {
+      "epoch": 0.43709669822429464,
+      "grad_norm": 2.403501510620117,
+      "learning_rate": 0.00011261363101125712,
+      "loss": 2.1115,
+      "step": 18560
+    },
+    {
+      "epoch": 0.4373322029108379,
+      "grad_norm": 2.2626802921295166,
+      "learning_rate": 0.00011256653007394849,
+      "loss": 2.089,
+      "step": 18570
+    },
+    {
+      "epoch": 0.4375677075973812,
+      "grad_norm": 2.3562400341033936,
+      "learning_rate": 0.00011251942913663982,
+      "loss": 2.3542,
+      "step": 18580
+    },
+    {
+      "epoch": 0.43780321228392444,
+      "grad_norm": 1.9294075965881348,
+      "learning_rate": 0.00011247232819933117,
+      "loss": 2.1626,
+      "step": 18590
+    },
+    {
+      "epoch": 0.4380387169704677,
+      "grad_norm": 2.3301680088043213,
+      "learning_rate": 0.00011242522726202251,
+      "loss": 2.047,
+      "step": 18600
+    },
+    {
+      "epoch": 0.43827422165701097,
+      "grad_norm": 2.25012469291687,
+      "learning_rate": 0.00011237812632471387,
+      "loss": 2.1893,
+      "step": 18610
+    },
+    {
+      "epoch": 0.43850972634355423,
+      "grad_norm": 2.4669597148895264,
+      "learning_rate": 0.00011233102538740521,
+      "loss": 1.9656,
+      "step": 18620
+    },
+    {
+      "epoch": 0.4387452310300975,
+      "grad_norm": 2.1987082958221436,
+      "learning_rate": 0.00011228392445009656,
+      "loss": 2.1176,
+      "step": 18630
+    },
+    {
+      "epoch": 0.43898073571664076,
+      "grad_norm": 2.573086738586426,
+      "learning_rate": 0.0001122368235127879,
+      "loss": 1.9486,
+      "step": 18640
+    },
+    {
+      "epoch": 0.439216240403184,
+      "grad_norm": 2.510371446609497,
+      "learning_rate": 0.00011218972257547926,
+      "loss": 2.2401,
+      "step": 18650
+    },
+    {
+      "epoch": 0.4394517450897273,
+      "grad_norm": 2.1042065620422363,
+      "learning_rate": 0.0001121426216381706,
+      "loss": 1.9531,
+      "step": 18660
+    },
+    {
+      "epoch": 0.43968724977627055,
+      "grad_norm": 2.6314985752105713,
+      "learning_rate": 0.00011209552070086196,
+      "loss": 2.126,
+      "step": 18670
+    },
+    {
+      "epoch": 0.4399227544628138,
+      "grad_norm": 1.911363959312439,
+      "learning_rate": 0.0001120484197635533,
+      "loss": 2.2783,
+      "step": 18680
+    },
+    {
+      "epoch": 0.4401582591493571,
+      "grad_norm": 3.6765568256378174,
+      "learning_rate": 0.00011200131882624465,
+      "loss": 2.0344,
+      "step": 18690
+    },
+    {
+      "epoch": 0.44039376383590034,
+      "grad_norm": 1.8842840194702148,
+      "learning_rate": 0.00011195421788893599,
+      "loss": 2.0042,
+      "step": 18700
+    },
+    {
+      "epoch": 0.4406292685224436,
+      "grad_norm": 2.389233112335205,
+      "learning_rate": 0.00011190711695162735,
+      "loss": 2.0097,
+      "step": 18710
+    },
+    {
+      "epoch": 0.44086477320898687,
+      "grad_norm": 1.9120182991027832,
+      "learning_rate": 0.00011186001601431869,
+      "loss": 1.9488,
+      "step": 18720
+    },
+    {
+      "epoch": 0.44110027789553014,
+      "grad_norm": 3.073310613632202,
+      "learning_rate": 0.00011181291507701004,
+      "loss": 1.8932,
+      "step": 18730
+    },
+    {
+      "epoch": 0.4413357825820734,
+      "grad_norm": 2.4336462020874023,
+      "learning_rate": 0.00011176581413970137,
+      "loss": 2.028,
+      "step": 18740
+    },
+    {
+      "epoch": 0.44157128726861666,
+      "grad_norm": 2.4928979873657227,
+      "learning_rate": 0.00011171871320239274,
+      "loss": 2.0821,
+      "step": 18750
+    },
+    {
+      "epoch": 0.4418067919551599,
+      "grad_norm": 2.694518804550171,
+      "learning_rate": 0.00011167161226508409,
+      "loss": 2.0103,
+      "step": 18760
+    },
+    {
+      "epoch": 0.4420422966417032,
+      "grad_norm": 1.9793258905410767,
+      "learning_rate": 0.00011162451132777542,
+      "loss": 2.0106,
+      "step": 18770
+    },
+    {
+      "epoch": 0.44227780132824646,
+      "grad_norm": 3.150414228439331,
+      "learning_rate": 0.00011157741039046679,
+      "loss": 2.2266,
+      "step": 18780
+    },
+    {
+      "epoch": 0.4425133060147897,
+      "grad_norm": 2.1798579692840576,
+      "learning_rate": 0.00011153030945315812,
+      "loss": 2.1064,
+      "step": 18790
+    },
+    {
+      "epoch": 0.44274881070133293,
+      "grad_norm": 2.7977168560028076,
+      "learning_rate": 0.00011148320851584947,
+      "loss": 2.179,
+      "step": 18800
+    },
+    {
+      "epoch": 0.4429843153878762,
+      "grad_norm": 1.93820059299469,
+      "learning_rate": 0.00011143610757854081,
+      "loss": 2.2933,
+      "step": 18810
+    },
+    {
+      "epoch": 0.44321982007441946,
+      "grad_norm": 2.5885021686553955,
+      "learning_rate": 0.00011138900664123217,
+      "loss": 2.1091,
+      "step": 18820
+    },
+    {
+      "epoch": 0.4434553247609627,
+      "grad_norm": 2.002413034439087,
+      "learning_rate": 0.00011134190570392351,
+      "loss": 1.8793,
+      "step": 18830
+    },
+    {
+      "epoch": 0.443690829447506,
+      "grad_norm": 2.016279697418213,
+      "learning_rate": 0.00011129480476661487,
+      "loss": 1.9843,
+      "step": 18840
+    },
+    {
+      "epoch": 0.44392633413404925,
+      "grad_norm": 1.719465732574463,
+      "learning_rate": 0.0001112477038293062,
+      "loss": 2.0675,
+      "step": 18850
+    },
+    {
+      "epoch": 0.4441618388205925,
+      "grad_norm": 2.9975037574768066,
+      "learning_rate": 0.00011120060289199756,
+      "loss": 1.9329,
+      "step": 18860
+    },
+    {
+      "epoch": 0.4443973435071358,
+      "grad_norm": 2.3108158111572266,
+      "learning_rate": 0.0001111535019546889,
+      "loss": 2.1686,
+      "step": 18870
+    },
+    {
+      "epoch": 0.44463284819367904,
+      "grad_norm": 2.196322202682495,
+      "learning_rate": 0.00011110640101738026,
+      "loss": 2.1436,
+      "step": 18880
+    },
+    {
+      "epoch": 0.4448683528802223,
+      "grad_norm": 2.0889739990234375,
+      "learning_rate": 0.0001110593000800716,
+      "loss": 2.1503,
+      "step": 18890
+    },
+    {
+      "epoch": 0.44510385756676557,
+      "grad_norm": 1.8750911951065063,
+      "learning_rate": 0.00011101219914276295,
+      "loss": 2.1539,
+      "step": 18900
+    },
+    {
+      "epoch": 0.44533936225330883,
+      "grad_norm": 2.5251245498657227,
+      "learning_rate": 0.00011096509820545429,
+      "loss": 2.0683,
+      "step": 18910
+    },
+    {
+      "epoch": 0.4455748669398521,
+      "grad_norm": 2.4985053539276123,
+      "learning_rate": 0.00011091799726814565,
+      "loss": 2.0526,
+      "step": 18920
+    },
+    {
+      "epoch": 0.44581037162639536,
+      "grad_norm": 2.0587680339813232,
+      "learning_rate": 0.00011087089633083699,
+      "loss": 2.1422,
+      "step": 18930
+    },
+    {
+      "epoch": 0.4460458763129386,
+      "grad_norm": 2.103192090988159,
+      "learning_rate": 0.00011082379539352834,
+      "loss": 1.9882,
+      "step": 18940
+    },
+    {
+      "epoch": 0.4462813809994819,
+      "grad_norm": 2.437361717224121,
+      "learning_rate": 0.00011077669445621967,
+      "loss": 2.1671,
+      "step": 18950
+    },
+    {
+      "epoch": 0.44651688568602516,
+      "grad_norm": 2.0805160999298096,
+      "learning_rate": 0.00011072959351891104,
+      "loss": 2.0893,
+      "step": 18960
+    },
+    {
+      "epoch": 0.4467523903725684,
+      "grad_norm": 2.183110475540161,
+      "learning_rate": 0.00011068249258160237,
+      "loss": 1.9785,
+      "step": 18970
+    },
+    {
+      "epoch": 0.4469878950591117,
+      "grad_norm": 3.0522234439849854,
+      "learning_rate": 0.00011063539164429372,
+      "loss": 2.1078,
+      "step": 18980
+    },
+    {
+      "epoch": 0.44722339974565495,
+      "grad_norm": 2.6682941913604736,
+      "learning_rate": 0.00011058829070698506,
+      "loss": 2.1685,
+      "step": 18990
+    },
+    {
+      "epoch": 0.4474589044321982,
+      "grad_norm": 2.421186685562134,
+      "learning_rate": 0.00011054118976967642,
+      "loss": 2.1228,
+      "step": 19000
+    },
+    {
+      "epoch": 0.4476944091187415,
+      "grad_norm": 1.8149057626724243,
+      "learning_rate": 0.00011049408883236776,
+      "loss": 2.1178,
+      "step": 19010
+    },
+    {
+      "epoch": 0.44792991380528474,
+      "grad_norm": 2.045279026031494,
+      "learning_rate": 0.00011044698789505911,
+      "loss": 1.934,
+      "step": 19020
+    },
+    {
+      "epoch": 0.448165418491828,
+      "grad_norm": 2.152418375015259,
+      "learning_rate": 0.00011039988695775045,
+      "loss": 2.0527,
+      "step": 19030
+    },
+    {
+      "epoch": 0.44840092317837127,
+      "grad_norm": 2.6455652713775635,
+      "learning_rate": 0.00011035278602044181,
+      "loss": 1.9045,
+      "step": 19040
+    },
+    {
+      "epoch": 0.44863642786491453,
+      "grad_norm": 1.9313437938690186,
+      "learning_rate": 0.00011030568508313318,
+      "loss": 2.0172,
+      "step": 19050
+    },
+    {
+      "epoch": 0.4488719325514578,
+      "grad_norm": 2.253701686859131,
+      "learning_rate": 0.00011025858414582451,
+      "loss": 2.164,
+      "step": 19060
+    },
+    {
+      "epoch": 0.44910743723800106,
+      "grad_norm": 1.9324719905853271,
+      "learning_rate": 0.00011021148320851586,
+      "loss": 2.333,
+      "step": 19070
+    },
+    {
+      "epoch": 0.4493429419245443,
+      "grad_norm": 1.903435230255127,
+      "learning_rate": 0.0001101643822712072,
+      "loss": 2.0204,
+      "step": 19080
+    },
+    {
+      "epoch": 0.44957844661108753,
+      "grad_norm": 2.1830732822418213,
+      "learning_rate": 0.00011011728133389856,
+      "loss": 2.2265,
+      "step": 19090
+    },
+    {
+      "epoch": 0.4498139512976308,
+      "grad_norm": 2.450608253479004,
+      "learning_rate": 0.0001100701803965899,
+      "loss": 2.047,
+      "step": 19100
+    },
+    {
+      "epoch": 0.45004945598417406,
+      "grad_norm": 2.1484196186065674,
+      "learning_rate": 0.00011002307945928125,
+      "loss": 1.9849,
+      "step": 19110
+    },
+    {
+      "epoch": 0.4502849606707173,
+      "grad_norm": 2.1579840183258057,
+      "learning_rate": 0.00010997597852197259,
+      "loss": 2.1648,
+      "step": 19120
+    },
+    {
+      "epoch": 0.4505204653572606,
+      "grad_norm": 2.8706860542297363,
+      "learning_rate": 0.00010992887758466395,
+      "loss": 1.9726,
+      "step": 19130
+    },
+    {
+      "epoch": 0.45075597004380386,
+      "grad_norm": 2.238607406616211,
+      "learning_rate": 0.00010988177664735529,
+      "loss": 2.056,
+      "step": 19140
+    },
+    {
+      "epoch": 0.4509914747303471,
+      "grad_norm": 1.9871670007705688,
+      "learning_rate": 0.00010983467571004664,
+      "loss": 2.1684,
+      "step": 19150
+    },
+    {
+      "epoch": 0.4512269794168904,
+      "grad_norm": 2.0434229373931885,
+      "learning_rate": 0.00010978757477273797,
+      "loss": 2.1818,
+      "step": 19160
+    },
+    {
+      "epoch": 0.45146248410343365,
+      "grad_norm": 1.8030943870544434,
+      "learning_rate": 0.00010974047383542934,
+      "loss": 1.9667,
+      "step": 19170
+    },
+    {
+      "epoch": 0.4516979887899769,
+      "grad_norm": 2.399859666824341,
+      "learning_rate": 0.00010969337289812067,
+      "loss": 1.9836,
+      "step": 19180
+    },
+    {
+      "epoch": 0.4519334934765202,
+      "grad_norm": 2.7550058364868164,
+      "learning_rate": 0.00010964627196081203,
+      "loss": 2.1151,
+      "step": 19190
+    },
+    {
+      "epoch": 0.45216899816306344,
+      "grad_norm": 2.3162660598754883,
+      "learning_rate": 0.00010959917102350336,
+      "loss": 2.0955,
+      "step": 19200
+    },
+    {
+      "epoch": 0.4524045028496067,
+      "grad_norm": 1.7464721202850342,
+      "learning_rate": 0.00010955678017992559,
+      "loss": 1.9926,
+      "step": 19210
+    },
+    {
+      "epoch": 0.45264000753614997,
+      "grad_norm": 3.1640796661376953,
+      "learning_rate": 0.00010950967924261693,
+      "loss": 2.0717,
+      "step": 19220
+    },
+    {
+      "epoch": 0.45287551222269323,
+      "grad_norm": 1.9667413234710693,
+      "learning_rate": 0.00010946257830530828,
+      "loss": 2.1224,
+      "step": 19230
+    },
+    {
+      "epoch": 0.4531110169092365,
+      "grad_norm": 2.176348924636841,
+      "learning_rate": 0.00010941547736799962,
+      "loss": 2.0679,
+      "step": 19240
+    },
+    {
+      "epoch": 0.45334652159577976,
+      "grad_norm": 2.226438045501709,
+      "learning_rate": 0.00010936837643069098,
+      "loss": 2.1582,
+      "step": 19250
+    },
+    {
+      "epoch": 0.453582026282323,
+      "grad_norm": 2.002314805984497,
+      "learning_rate": 0.00010932127549338232,
+      "loss": 2.0523,
+      "step": 19260
+    },
+    {
+      "epoch": 0.4538175309688663,
+      "grad_norm": 3.0285584926605225,
+      "learning_rate": 0.00010927417455607367,
+      "loss": 2.1907,
+      "step": 19270
+    },
+    {
+      "epoch": 0.45405303565540955,
+      "grad_norm": 3.506699323654175,
+      "learning_rate": 0.000109227073618765,
+      "loss": 2.1037,
+      "step": 19280
+    },
+    {
+      "epoch": 0.4542885403419528,
+      "grad_norm": 2.3388898372650146,
+      "learning_rate": 0.00010917997268145637,
+      "loss": 2.0536,
+      "step": 19290
+    },
+    {
+      "epoch": 0.4545240450284961,
+      "grad_norm": 1.9860022068023682,
+      "learning_rate": 0.00010913287174414772,
+      "loss": 2.1253,
+      "step": 19300
+    },
+    {
+      "epoch": 0.45475954971503935,
+      "grad_norm": 2.931128740310669,
+      "learning_rate": 0.00010908577080683905,
+      "loss": 2.1019,
+      "step": 19310
+    },
+    {
+      "epoch": 0.4549950544015826,
+      "grad_norm": 1.9610087871551514,
+      "learning_rate": 0.00010903866986953042,
+      "loss": 2.114,
+      "step": 19320
+    },
+    {
+      "epoch": 0.4552305590881259,
+      "grad_norm": 2.159877300262451,
+      "learning_rate": 0.00010899156893222176,
+      "loss": 1.8286,
+      "step": 19330
+    },
+    {
+      "epoch": 0.45546606377466914,
+      "grad_norm": 1.8019903898239136,
+      "learning_rate": 0.00010894446799491312,
+      "loss": 2.1,
+      "step": 19340
+    },
+    {
+      "epoch": 0.4557015684612124,
+      "grad_norm": 2.368715524673462,
+      "learning_rate": 0.00010889736705760444,
+      "loss": 2.1636,
+      "step": 19350
+    },
+    {
+      "epoch": 0.45593707314775567,
+      "grad_norm": 2.2068965435028076,
+      "learning_rate": 0.0001088502661202958,
+      "loss": 2.1377,
+      "step": 19360
+    },
+    {
+      "epoch": 0.4561725778342989,
+      "grad_norm": 2.2656490802764893,
+      "learning_rate": 0.00010880316518298714,
+      "loss": 2.1331,
+      "step": 19370
+    },
+    {
+      "epoch": 0.45640808252084214,
+      "grad_norm": 1.6999800205230713,
+      "learning_rate": 0.0001087560642456785,
+      "loss": 1.8618,
+      "step": 19380
+    },
+    {
+      "epoch": 0.4566435872073854,
+      "grad_norm": 2.0297725200653076,
+      "learning_rate": 0.00010870896330836984,
+      "loss": 2.0886,
+      "step": 19390
+    },
+    {
+      "epoch": 0.45687909189392867,
+      "grad_norm": 2.10459566116333,
+      "learning_rate": 0.0001086618623710612,
+      "loss": 2.0182,
+      "step": 19400
+    },
+    {
+      "epoch": 0.45711459658047193,
+      "grad_norm": 2.357388496398926,
+      "learning_rate": 0.00010861476143375253,
+      "loss": 1.9764,
+      "step": 19410
+    },
+    {
+      "epoch": 0.4573501012670152,
+      "grad_norm": 2.0262765884399414,
+      "learning_rate": 0.0001085676604964439,
+      "loss": 2.3218,
+      "step": 19420
+    },
+    {
+      "epoch": 0.45758560595355846,
+      "grad_norm": 3.8287534713745117,
+      "learning_rate": 0.00010852055955913523,
+      "loss": 2.1722,
+      "step": 19430
+    },
+    {
+      "epoch": 0.4578211106401017,
+      "grad_norm": 2.3780245780944824,
+      "learning_rate": 0.00010847345862182658,
+      "loss": 2.0486,
+      "step": 19440
+    },
+    {
+      "epoch": 0.458056615326645,
+      "grad_norm": 2.3219287395477295,
+      "learning_rate": 0.00010842635768451792,
+      "loss": 2.2132,
+      "step": 19450
+    },
+    {
+      "epoch": 0.45829212001318825,
+      "grad_norm": 2.278029680252075,
+      "learning_rate": 0.00010837925674720928,
+      "loss": 1.9851,
+      "step": 19460
+    },
+    {
+      "epoch": 0.4585276246997315,
+      "grad_norm": 2.079396963119507,
+      "learning_rate": 0.00010833215580990062,
+      "loss": 2.2242,
+      "step": 19470
+    },
+    {
+      "epoch": 0.4587631293862748,
+      "grad_norm": 2.292618751525879,
+      "learning_rate": 0.00010828505487259197,
+      "loss": 1.912,
+      "step": 19480
+    },
+    {
+      "epoch": 0.45899863407281805,
+      "grad_norm": 2.677804708480835,
+      "learning_rate": 0.0001082379539352833,
+      "loss": 2.1219,
+      "step": 19490
+    },
+    {
+      "epoch": 0.4592341387593613,
+      "grad_norm": 2.149505138397217,
+      "learning_rate": 0.00010819085299797467,
+      "loss": 2.0128,
+      "step": 19500
+    },
+    {
+      "epoch": 0.4594696434459046,
+      "grad_norm": 2.596064329147339,
+      "learning_rate": 0.000108143752060666,
+      "loss": 2.1132,
+      "step": 19510
+    },
+    {
+      "epoch": 0.45970514813244784,
+      "grad_norm": 2.409451484680176,
+      "learning_rate": 0.00010809665112335736,
+      "loss": 1.9036,
+      "step": 19520
+    },
+    {
+      "epoch": 0.4599406528189911,
+      "grad_norm": 2.5537643432617188,
+      "learning_rate": 0.00010804955018604869,
+      "loss": 1.9008,
+      "step": 19530
+    },
+    {
+      "epoch": 0.46017615750553437,
+      "grad_norm": 1.7899531126022339,
+      "learning_rate": 0.00010800244924874006,
+      "loss": 1.829,
+      "step": 19540
+    },
+    {
+      "epoch": 0.46041166219207763,
+      "grad_norm": 1.7132242918014526,
+      "learning_rate": 0.00010795534831143139,
+      "loss": 2.1235,
+      "step": 19550
+    },
+    {
+      "epoch": 0.4606471668786209,
+      "grad_norm": 2.0346591472625732,
+      "learning_rate": 0.00010790824737412276,
+      "loss": 2.0902,
+      "step": 19560
+    },
+    {
+      "epoch": 0.46088267156516416,
+      "grad_norm": 3.4156370162963867,
+      "learning_rate": 0.00010786114643681408,
+      "loss": 2.2832,
+      "step": 19570
+    },
+    {
+      "epoch": 0.4611181762517074,
+      "grad_norm": 2.067667007446289,
+      "learning_rate": 0.00010781404549950544,
+      "loss": 2.1939,
+      "step": 19580
+    },
+    {
+      "epoch": 0.4613536809382507,
+      "grad_norm": 1.9971282482147217,
+      "learning_rate": 0.00010776694456219681,
+      "loss": 2.1049,
+      "step": 19590
+    },
+    {
+      "epoch": 0.46158918562479395,
+      "grad_norm": 2.331735849380493,
+      "learning_rate": 0.00010771984362488814,
+      "loss": 2.3262,
+      "step": 19600
+    },
+    {
+      "epoch": 0.4618246903113372,
+      "grad_norm": 2.4964492321014404,
+      "learning_rate": 0.0001076727426875795,
+      "loss": 1.8584,
+      "step": 19610
+    },
+    {
+      "epoch": 0.4620601949978805,
+      "grad_norm": 2.145075559616089,
+      "learning_rate": 0.00010762564175027083,
+      "loss": 2.1807,
+      "step": 19620
+    },
+    {
+      "epoch": 0.46229569968442374,
+      "grad_norm": 2.1231913566589355,
+      "learning_rate": 0.00010758325090669306,
+      "loss": 2.0905,
+      "step": 19630
+    },
+    {
+      "epoch": 0.462531204370967,
+      "grad_norm": 2.6678659915924072,
+      "learning_rate": 0.00010753614996938439,
+      "loss": 1.9073,
+      "step": 19640
+    },
+    {
+      "epoch": 0.4627667090575102,
+      "grad_norm": 1.718123435974121,
+      "learning_rate": 0.00010748904903207575,
+      "loss": 2.1715,
+      "step": 19650
+    },
+    {
+      "epoch": 0.4630022137440535,
+      "grad_norm": 2.190896511077881,
+      "learning_rate": 0.00010744194809476709,
+      "loss": 2.1435,
+      "step": 19660
+    },
+    {
+      "epoch": 0.46323771843059675,
+      "grad_norm": 2.8094241619110107,
+      "learning_rate": 0.00010739484715745845,
+      "loss": 1.9762,
+      "step": 19670
+    },
+    {
+      "epoch": 0.46347322311714,
+      "grad_norm": 1.8374676704406738,
+      "learning_rate": 0.00010734774622014979,
+      "loss": 2.0374,
+      "step": 19680
+    },
+    {
+      "epoch": 0.4637087278036833,
+      "grad_norm": 2.1029560565948486,
+      "learning_rate": 0.00010730064528284114,
+      "loss": 2.3126,
+      "step": 19690
+    },
+    {
+      "epoch": 0.46394423249022654,
+      "grad_norm": 2.3267154693603516,
+      "learning_rate": 0.00010725354434553247,
+      "loss": 2.0234,
+      "step": 19700
+    },
+    {
+      "epoch": 0.4641797371767698,
+      "grad_norm": 2.216888189315796,
+      "learning_rate": 0.00010720644340822384,
+      "loss": 2.0581,
+      "step": 19710
+    },
+    {
+      "epoch": 0.46441524186331307,
+      "grad_norm": 2.1676642894744873,
+      "learning_rate": 0.00010715934247091517,
+      "loss": 2.0509,
+      "step": 19720
+    },
+    {
+      "epoch": 0.46465074654985633,
+      "grad_norm": 2.206120014190674,
+      "learning_rate": 0.00010711224153360652,
+      "loss": 2.0382,
+      "step": 19730
+    },
+    {
+      "epoch": 0.4648862512363996,
+      "grad_norm": 2.1404688358306885,
+      "learning_rate": 0.00010706514059629786,
+      "loss": 2.086,
+      "step": 19740
+    },
+    {
+      "epoch": 0.46512175592294286,
+      "grad_norm": 2.2367289066314697,
+      "learning_rate": 0.00010701803965898922,
+      "loss": 2.2163,
+      "step": 19750
+    },
+    {
+      "epoch": 0.4653572606094861,
+      "grad_norm": 1.8468304872512817,
+      "learning_rate": 0.00010697093872168056,
+      "loss": 2.1602,
+      "step": 19760
+    },
+    {
+      "epoch": 0.4655927652960294,
+      "grad_norm": 2.4168541431427,
+      "learning_rate": 0.00010692383778437191,
+      "loss": 2.0497,
+      "step": 19770
+    },
+    {
+      "epoch": 0.46582826998257265,
+      "grad_norm": 1.9554232358932495,
+      "learning_rate": 0.00010687673684706325,
+      "loss": 2.1027,
+      "step": 19780
+    },
+    {
+      "epoch": 0.4660637746691159,
+      "grad_norm": 2.9504382610321045,
+      "learning_rate": 0.00010682963590975461,
+      "loss": 1.9,
+      "step": 19790
+    },
+    {
+      "epoch": 0.4662992793556592,
+      "grad_norm": 3.071281671524048,
+      "learning_rate": 0.00010678253497244595,
+      "loss": 2.0025,
+      "step": 19800
+    },
+    {
+      "epoch": 0.46653478404220244,
+      "grad_norm": 2.1430885791778564,
+      "learning_rate": 0.0001067354340351373,
+      "loss": 1.8642,
+      "step": 19810
+    },
+    {
+      "epoch": 0.4667702887287457,
+      "grad_norm": 1.6342874765396118,
+      "learning_rate": 0.00010668833309782864,
+      "loss": 2.0941,
+      "step": 19820
+    },
+    {
+      "epoch": 0.467005793415289,
+      "grad_norm": 1.8183525800704956,
+      "learning_rate": 0.00010664123216052,
+      "loss": 2.0612,
+      "step": 19830
+    },
+    {
+      "epoch": 0.46724129810183224,
+      "grad_norm": 2.2085859775543213,
+      "learning_rate": 0.00010659413122321136,
+      "loss": 2.0461,
+      "step": 19840
+    },
+    {
+      "epoch": 0.4674768027883755,
+      "grad_norm": 2.12282395362854,
+      "learning_rate": 0.0001065470302859027,
+      "loss": 2.0324,
+      "step": 19850
+    },
+    {
+      "epoch": 0.46771230747491876,
+      "grad_norm": 2.824951648712158,
+      "learning_rate": 0.00010649992934859405,
+      "loss": 1.9978,
+      "step": 19860
+    },
+    {
+      "epoch": 0.46794781216146203,
+      "grad_norm": 1.9392645359039307,
+      "learning_rate": 0.00010645282841128539,
+      "loss": 2.2445,
+      "step": 19870
+    },
+    {
+      "epoch": 0.4681833168480053,
+      "grad_norm": 2.1611478328704834,
+      "learning_rate": 0.00010640572747397675,
+      "loss": 2.1556,
+      "step": 19880
+    },
+    {
+      "epoch": 0.46841882153454856,
+      "grad_norm": 2.281257390975952,
+      "learning_rate": 0.00010635862653666809,
+      "loss": 2.0552,
+      "step": 19890
+    },
+    {
+      "epoch": 0.4686543262210918,
+      "grad_norm": 2.2869932651519775,
+      "learning_rate": 0.00010631152559935944,
+      "loss": 2.1354,
+      "step": 19900
+    },
+    {
+      "epoch": 0.4688898309076351,
+      "grad_norm": 2.583472967147827,
+      "learning_rate": 0.00010626442466205077,
+      "loss": 2.0894,
+      "step": 19910
+    },
+    {
+      "epoch": 0.46912533559417835,
+      "grad_norm": 1.9615871906280518,
+      "learning_rate": 0.00010621732372474214,
+      "loss": 2.0168,
+      "step": 19920
+    },
+    {
+      "epoch": 0.4693608402807216,
+      "grad_norm": 2.0687925815582275,
+      "learning_rate": 0.00010617022278743348,
+      "loss": 1.9977,
+      "step": 19930
+    },
+    {
+      "epoch": 0.4695963449672648,
+      "grad_norm": 1.8753052949905396,
+      "learning_rate": 0.00010612312185012483,
+      "loss": 1.9732,
+      "step": 19940
+    },
+    {
+      "epoch": 0.4698318496538081,
+      "grad_norm": 2.3094046115875244,
+      "learning_rate": 0.00010607602091281616,
+      "loss": 2.1751,
+      "step": 19950
+    },
+    {
+      "epoch": 0.47006735434035135,
+      "grad_norm": 1.8566091060638428,
+      "learning_rate": 0.00010602891997550753,
+      "loss": 2.097,
+      "step": 19960
+    },
+    {
+      "epoch": 0.4703028590268946,
+      "grad_norm": 2.8540568351745605,
+      "learning_rate": 0.00010598181903819886,
+      "loss": 2.0623,
+      "step": 19970
+    },
+    {
+      "epoch": 0.4705383637134379,
+      "grad_norm": 2.5797126293182373,
+      "learning_rate": 0.00010593471810089021,
+      "loss": 2.0501,
+      "step": 19980
+    },
+    {
+      "epoch": 0.47077386839998114,
+      "grad_norm": 2.1992506980895996,
+      "learning_rate": 0.00010588761716358155,
+      "loss": 1.9637,
+      "step": 19990
+    },
+    {
+      "epoch": 0.4710093730865244,
+      "grad_norm": 1.7316781282424927,
+      "learning_rate": 0.00010584051622627291,
+      "loss": 2.156,
+      "step": 20000
+    },
+    {
+      "epoch": 0.47124487777306767,
+      "grad_norm": 2.292141914367676,
+      "learning_rate": 0.00010579341528896425,
+      "loss": 2.1193,
+      "step": 20010
+    },
+    {
+      "epoch": 0.47148038245961094,
+      "grad_norm": 2.6842830181121826,
+      "learning_rate": 0.0001057463143516556,
+      "loss": 2.0507,
+      "step": 20020
+    },
+    {
+      "epoch": 0.4717158871461542,
+      "grad_norm": 1.8671191930770874,
+      "learning_rate": 0.00010569921341434694,
+      "loss": 1.8824,
+      "step": 20030
+    },
+    {
+      "epoch": 0.47195139183269746,
+      "grad_norm": 2.062864303588867,
+      "learning_rate": 0.0001056521124770383,
+      "loss": 2.1032,
+      "step": 20040
+    },
+    {
+      "epoch": 0.47218689651924073,
+      "grad_norm": 1.6221140623092651,
+      "learning_rate": 0.00010560501153972964,
+      "loss": 2.2215,
+      "step": 20050
+    },
+    {
+      "epoch": 0.472422401205784,
+      "grad_norm": 3.1304547786712646,
+      "learning_rate": 0.000105557910602421,
+      "loss": 2.106,
+      "step": 20060
+    },
+    {
+      "epoch": 0.47265790589232726,
+      "grad_norm": 2.085540533065796,
+      "learning_rate": 0.00010551080966511234,
+      "loss": 2.1274,
+      "step": 20070
+    },
+    {
+      "epoch": 0.4728934105788705,
+      "grad_norm": 1.9874452352523804,
+      "learning_rate": 0.00010546370872780369,
+      "loss": 1.8995,
+      "step": 20080
+    },
+    {
+      "epoch": 0.4731289152654138,
+      "grad_norm": 2.646638870239258,
+      "learning_rate": 0.00010541660779049502,
+      "loss": 2.1751,
+      "step": 20090
+    },
+    {
+      "epoch": 0.47336441995195705,
+      "grad_norm": 2.6359567642211914,
+      "learning_rate": 0.00010536950685318639,
+      "loss": 1.9397,
+      "step": 20100
+    },
+    {
+      "epoch": 0.4735999246385003,
+      "grad_norm": 2.21799635887146,
+      "learning_rate": 0.00010532240591587774,
+      "loss": 2.1049,
+      "step": 20110
+    },
+    {
+      "epoch": 0.4738354293250436,
+      "grad_norm": 2.5067179203033447,
+      "learning_rate": 0.00010527530497856908,
+      "loss": 2.174,
+      "step": 20120
+    },
+    {
+      "epoch": 0.47407093401158684,
+      "grad_norm": 2.724372148513794,
+      "learning_rate": 0.00010522820404126044,
+      "loss": 2.1434,
+      "step": 20130
+    },
+    {
+      "epoch": 0.4743064386981301,
+      "grad_norm": 2.247986078262329,
+      "learning_rate": 0.00010518110310395178,
+      "loss": 2.0298,
+      "step": 20140
+    },
+    {
+      "epoch": 0.47454194338467337,
+      "grad_norm": 1.7842614650726318,
+      "learning_rate": 0.00010513400216664313,
+      "loss": 2.0235,
+      "step": 20150
+    },
+    {
+      "epoch": 0.47477744807121663,
+      "grad_norm": 2.239865303039551,
+      "learning_rate": 0.00010508690122933446,
+      "loss": 2.0488,
+      "step": 20160
+    },
+    {
+      "epoch": 0.4750129527577599,
+      "grad_norm": 1.8381290435791016,
+      "learning_rate": 0.00010503980029202583,
+      "loss": 2.137,
+      "step": 20170
+    },
+    {
+      "epoch": 0.47524845744430316,
+      "grad_norm": 1.837793231010437,
+      "learning_rate": 0.00010499269935471716,
+      "loss": 1.7239,
+      "step": 20180
+    },
+    {
+      "epoch": 0.4754839621308464,
+      "grad_norm": 2.2856781482696533,
+      "learning_rate": 0.00010494559841740851,
+      "loss": 2.0022,
+      "step": 20190
+    },
+    {
+      "epoch": 0.4757194668173897,
+      "grad_norm": 2.448535919189453,
+      "learning_rate": 0.00010489849748009985,
+      "loss": 2.191,
+      "step": 20200
+    },
+    {
+      "epoch": 0.47595497150393296,
+      "grad_norm": 2.522270679473877,
+      "learning_rate": 0.00010485139654279121,
+      "loss": 2.1016,
+      "step": 20210
+    },
+    {
+      "epoch": 0.47619047619047616,
+      "grad_norm": 2.3444366455078125,
+      "learning_rate": 0.00010480429560548255,
+      "loss": 2.1629,
+      "step": 20220
+    },
+    {
+      "epoch": 0.47642598087701943,
+      "grad_norm": 1.8811753988265991,
+      "learning_rate": 0.00010475719466817391,
+      "loss": 1.9501,
+      "step": 20230
+    },
+    {
+      "epoch": 0.4766614855635627,
+      "grad_norm": 2.2614917755126953,
+      "learning_rate": 0.00010471009373086524,
+      "loss": 1.9971,
+      "step": 20240
+    },
+    {
+      "epoch": 0.47689699025010596,
+      "grad_norm": 2.2525382041931152,
+      "learning_rate": 0.0001046629927935566,
+      "loss": 2.0558,
+      "step": 20250
+    },
+    {
+      "epoch": 0.4771324949366492,
+      "grad_norm": 3.0018701553344727,
+      "learning_rate": 0.00010461589185624794,
+      "loss": 2.1558,
+      "step": 20260
+    },
+    {
+      "epoch": 0.4773679996231925,
+      "grad_norm": 2.163959264755249,
+      "learning_rate": 0.0001045687909189393,
+      "loss": 2.0081,
+      "step": 20270
+    },
+    {
+      "epoch": 0.47760350430973575,
+      "grad_norm": 1.9860162734985352,
+      "learning_rate": 0.00010452168998163064,
+      "loss": 2.0474,
+      "step": 20280
+    },
+    {
+      "epoch": 0.477839008996279,
+      "grad_norm": 2.7075717449188232,
+      "learning_rate": 0.00010447458904432199,
+      "loss": 2.0118,
+      "step": 20290
+    },
+    {
+      "epoch": 0.4780745136828223,
+      "grad_norm": 2.2994916439056396,
+      "learning_rate": 0.00010442748810701333,
+      "loss": 1.9586,
+      "step": 20300
+    },
+    {
+      "epoch": 0.47831001836936554,
+      "grad_norm": 1.9415532350540161,
+      "learning_rate": 0.00010438038716970469,
+      "loss": 2.1362,
+      "step": 20310
+    },
+    {
+      "epoch": 0.4785455230559088,
+      "grad_norm": 2.319530725479126,
+      "learning_rate": 0.00010433328623239603,
+      "loss": 1.8642,
+      "step": 20320
+    },
+    {
+      "epoch": 0.47878102774245207,
+      "grad_norm": 2.8633079528808594,
+      "learning_rate": 0.00010428618529508738,
+      "loss": 2.2701,
+      "step": 20330
+    },
+    {
+      "epoch": 0.47901653242899533,
+      "grad_norm": 3.3309438228607178,
+      "learning_rate": 0.00010423908435777871,
+      "loss": 2.0667,
+      "step": 20340
+    },
+    {
+      "epoch": 0.4792520371155386,
+      "grad_norm": 1.8573964834213257,
+      "learning_rate": 0.00010419198342047008,
+      "loss": 2.1726,
+      "step": 20350
+    },
+    {
+      "epoch": 0.47948754180208186,
+      "grad_norm": 2.2213289737701416,
+      "learning_rate": 0.00010414488248316141,
+      "loss": 2.1585,
+      "step": 20360
+    },
+    {
+      "epoch": 0.4797230464886251,
+      "grad_norm": 2.0454204082489014,
+      "learning_rate": 0.00010409778154585276,
+      "loss": 2.1965,
+      "step": 20370
+    },
+    {
+      "epoch": 0.4799585511751684,
+      "grad_norm": 2.0503110885620117,
+      "learning_rate": 0.0001040506806085441,
+      "loss": 2.0648,
+      "step": 20380
+    },
+    {
+      "epoch": 0.48019405586171166,
+      "grad_norm": 2.948557138442993,
+      "learning_rate": 0.00010400357967123546,
+      "loss": 1.9269,
+      "step": 20390
+    },
+    {
+      "epoch": 0.4804295605482549,
+      "grad_norm": 2.1079161167144775,
+      "learning_rate": 0.00010395647873392681,
+      "loss": 1.9769,
+      "step": 20400
+    },
+    {
+      "epoch": 0.4806650652347982,
+      "grad_norm": 2.086078643798828,
+      "learning_rate": 0.00010390937779661815,
+      "loss": 2.0752,
+      "step": 20410
+    },
+    {
+      "epoch": 0.48090056992134145,
+      "grad_norm": 1.7482393980026245,
+      "learning_rate": 0.00010386227685930952,
+      "loss": 1.9872,
+      "step": 20420
+    },
+    {
+      "epoch": 0.4811360746078847,
+      "grad_norm": 2.2785568237304688,
+      "learning_rate": 0.00010381517592200085,
+      "loss": 2.1667,
+      "step": 20430
+    },
+    {
+      "epoch": 0.481371579294428,
+      "grad_norm": 2.2354800701141357,
+      "learning_rate": 0.00010376807498469222,
+      "loss": 2.0142,
+      "step": 20440
+    },
+    {
+      "epoch": 0.48160708398097124,
+      "grad_norm": 2.2600879669189453,
+      "learning_rate": 0.00010372097404738355,
+      "loss": 2.1101,
+      "step": 20450
+    },
+    {
+      "epoch": 0.4818425886675145,
+      "grad_norm": 2.2491447925567627,
+      "learning_rate": 0.0001036738731100749,
+      "loss": 2.0255,
+      "step": 20460
+    },
+    {
+      "epoch": 0.48207809335405777,
+      "grad_norm": 3.314981698989868,
+      "learning_rate": 0.00010362677217276624,
+      "loss": 1.8906,
+      "step": 20470
+    },
+    {
+      "epoch": 0.48231359804060103,
+      "grad_norm": 2.204580783843994,
+      "learning_rate": 0.0001035796712354576,
+      "loss": 2.235,
+      "step": 20480
+    },
+    {
+      "epoch": 0.4825491027271443,
+      "grad_norm": 2.513540029525757,
+      "learning_rate": 0.00010353257029814894,
+      "loss": 1.9714,
+      "step": 20490
+    },
+    {
+      "epoch": 0.4827846074136875,
+      "grad_norm": 2.1390328407287598,
+      "learning_rate": 0.00010348546936084029,
+      "loss": 2.242,
+      "step": 20500
+    },
+    {
+      "epoch": 0.48302011210023077,
+      "grad_norm": 1.9267793893814087,
+      "learning_rate": 0.00010343836842353163,
+      "loss": 2.2083,
+      "step": 20510
+    },
+    {
+      "epoch": 0.48325561678677403,
+      "grad_norm": 1.821351408958435,
+      "learning_rate": 0.00010339126748622299,
+      "loss": 2.0652,
+      "step": 20520
+    },
+    {
+      "epoch": 0.4834911214733173,
+      "grad_norm": 2.3727545738220215,
+      "learning_rate": 0.00010334416654891433,
+      "loss": 2.2172,
+      "step": 20530
+    },
+    {
+      "epoch": 0.48372662615986056,
+      "grad_norm": 2.450948476791382,
+      "learning_rate": 0.00010329706561160568,
+      "loss": 2.0812,
+      "step": 20540
+    },
+    {
+      "epoch": 0.4839621308464038,
+      "grad_norm": 2.104326009750366,
+      "learning_rate": 0.00010324996467429701,
+      "loss": 2.1296,
+      "step": 20550
+    },
+    {
+      "epoch": 0.4841976355329471,
+      "grad_norm": 1.6081938743591309,
+      "learning_rate": 0.00010320286373698838,
+      "loss": 2.1119,
+      "step": 20560
+    },
+    {
+      "epoch": 0.48443314021949035,
+      "grad_norm": 2.259737491607666,
+      "learning_rate": 0.00010315576279967971,
+      "loss": 2.0855,
+      "step": 20570
+    },
+    {
+      "epoch": 0.4846686449060336,
+      "grad_norm": 2.6324872970581055,
+      "learning_rate": 0.00010310866186237106,
+      "loss": 2.1031,
+      "step": 20580
+    },
+    {
+      "epoch": 0.4849041495925769,
+      "grad_norm": 2.676846504211426,
+      "learning_rate": 0.0001030615609250624,
+      "loss": 1.8389,
+      "step": 20590
+    },
+    {
+      "epoch": 0.48513965427912015,
+      "grad_norm": 2.062814712524414,
+      "learning_rate": 0.00010301445998775377,
+      "loss": 2.0708,
+      "step": 20600
+    },
+    {
+      "epoch": 0.4853751589656634,
+      "grad_norm": 1.8760517835617065,
+      "learning_rate": 0.0001029673590504451,
+      "loss": 1.9559,
+      "step": 20610
+    },
+    {
+      "epoch": 0.4856106636522067,
+      "grad_norm": 2.316432476043701,
+      "learning_rate": 0.00010292025811313645,
+      "loss": 2.0927,
+      "step": 20620
+    },
+    {
+      "epoch": 0.48584616833874994,
+      "grad_norm": 2.2129321098327637,
+      "learning_rate": 0.00010287315717582779,
+      "loss": 2.0794,
+      "step": 20630
+    },
+    {
+      "epoch": 0.4860816730252932,
+      "grad_norm": 2.3312692642211914,
+      "learning_rate": 0.00010282605623851915,
+      "loss": 1.9387,
+      "step": 20640
+    },
+    {
+      "epoch": 0.48631717771183647,
+      "grad_norm": 1.807623267173767,
+      "learning_rate": 0.00010277895530121049,
+      "loss": 1.984,
+      "step": 20650
+    },
+    {
+      "epoch": 0.48655268239837973,
+      "grad_norm": 2.3233418464660645,
+      "learning_rate": 0.00010273185436390185,
+      "loss": 1.8548,
+      "step": 20660
+    },
+    {
+      "epoch": 0.486788187084923,
+      "grad_norm": 2.37127423286438,
+      "learning_rate": 0.00010268475342659319,
+      "loss": 2.2169,
+      "step": 20670
+    },
+    {
+      "epoch": 0.48702369177146626,
+      "grad_norm": 2.1091103553771973,
+      "learning_rate": 0.00010263765248928454,
+      "loss": 2.0467,
+      "step": 20680
+    },
+    {
+      "epoch": 0.4872591964580095,
+      "grad_norm": 3.436962127685547,
+      "learning_rate": 0.0001025905515519759,
+      "loss": 2.1332,
+      "step": 20690
+    },
+    {
+      "epoch": 0.4874947011445528,
+      "grad_norm": 2.2271640300750732,
+      "learning_rate": 0.00010254345061466724,
+      "loss": 2.0451,
+      "step": 20700
+    },
+    {
+      "epoch": 0.48773020583109605,
+      "grad_norm": 2.0807104110717773,
+      "learning_rate": 0.00010249634967735859,
+      "loss": 1.9054,
+      "step": 20710
+    },
+    {
+      "epoch": 0.4879657105176393,
+      "grad_norm": 2.3256382942199707,
+      "learning_rate": 0.00010244924874004993,
+      "loss": 1.9937,
+      "step": 20720
+    },
+    {
+      "epoch": 0.4882012152041826,
+      "grad_norm": 1.8084198236465454,
+      "learning_rate": 0.00010240214780274129,
+      "loss": 2.0964,
+      "step": 20730
+    },
+    {
+      "epoch": 0.48843671989072585,
+      "grad_norm": 2.9819064140319824,
+      "learning_rate": 0.00010235504686543263,
+      "loss": 2.0176,
+      "step": 20740
+    },
+    {
+      "epoch": 0.4886722245772691,
+      "grad_norm": 2.723691701889038,
+      "learning_rate": 0.00010230794592812398,
+      "loss": 2.2222,
+      "step": 20750
+    },
+    {
+      "epoch": 0.4889077292638124,
+      "grad_norm": 2.697124719619751,
+      "learning_rate": 0.00010226084499081532,
+      "loss": 2.1921,
+      "step": 20760
+    },
+    {
+      "epoch": 0.48914323395035564,
+      "grad_norm": 2.0441181659698486,
+      "learning_rate": 0.00010221374405350668,
+      "loss": 2.1415,
+      "step": 20770
+    },
+    {
+      "epoch": 0.48937873863689885,
+      "grad_norm": 2.7658491134643555,
+      "learning_rate": 0.00010216664311619802,
+      "loss": 2.068,
+      "step": 20780
+    },
+    {
+      "epoch": 0.4896142433234421,
+      "grad_norm": 2.291435718536377,
+      "learning_rate": 0.00010211954217888937,
+      "loss": 1.9272,
+      "step": 20790
+    },
+    {
+      "epoch": 0.4898497480099854,
+      "grad_norm": 2.1761622428894043,
+      "learning_rate": 0.0001020724412415807,
+      "loss": 2.1504,
+      "step": 20800
+    },
+    {
+      "epoch": 0.49008525269652864,
+      "grad_norm": 1.9706686735153198,
+      "learning_rate": 0.00010202534030427207,
+      "loss": 2.1641,
+      "step": 20810
+    },
+    {
+      "epoch": 0.4903207573830719,
+      "grad_norm": 2.0886855125427246,
+      "learning_rate": 0.0001019782393669634,
+      "loss": 2.0652,
+      "step": 20820
+    },
+    {
+      "epoch": 0.49055626206961517,
+      "grad_norm": 2.300699234008789,
+      "learning_rate": 0.00010193113842965477,
+      "loss": 1.9582,
+      "step": 20830
+    },
+    {
+      "epoch": 0.49079176675615843,
+      "grad_norm": 2.276219606399536,
+      "learning_rate": 0.00010188403749234609,
+      "loss": 2.3446,
+      "step": 20840
+    },
+    {
+      "epoch": 0.4910272714427017,
+      "grad_norm": 1.8282259702682495,
+      "learning_rate": 0.00010183693655503745,
+      "loss": 2.1345,
+      "step": 20850
+    },
+    {
+      "epoch": 0.49126277612924496,
+      "grad_norm": 2.4664013385772705,
+      "learning_rate": 0.00010178983561772879,
+      "loss": 2.081,
+      "step": 20860
+    },
+    {
+      "epoch": 0.4914982808157882,
+      "grad_norm": 2.2036657333374023,
+      "learning_rate": 0.00010174273468042015,
+      "loss": 1.9418,
+      "step": 20870
+    },
+    {
+      "epoch": 0.4917337855023315,
+      "grad_norm": 2.229975938796997,
+      "learning_rate": 0.00010169563374311149,
+      "loss": 2.0532,
+      "step": 20880
+    },
+    {
+      "epoch": 0.49196929018887475,
+      "grad_norm": 2.0748679637908936,
+      "learning_rate": 0.00010164853280580284,
+      "loss": 2.0162,
+      "step": 20890
+    },
+    {
+      "epoch": 0.492204794875418,
+      "grad_norm": 2.1549928188323975,
+      "learning_rate": 0.00010160143186849418,
+      "loss": 2.1265,
+      "step": 20900
+    },
+    {
+      "epoch": 0.4924402995619613,
+      "grad_norm": 2.1036789417266846,
+      "learning_rate": 0.00010155433093118554,
+      "loss": 2.1916,
+      "step": 20910
+    },
+    {
+      "epoch": 0.49267580424850455,
+      "grad_norm": 1.9950534105300903,
+      "learning_rate": 0.00010150722999387688,
+      "loss": 1.9724,
+      "step": 20920
+    },
+    {
+      "epoch": 0.4929113089350478,
+      "grad_norm": 1.8147923946380615,
+      "learning_rate": 0.00010146012905656823,
+      "loss": 1.9993,
+      "step": 20930
+    },
+    {
+      "epoch": 0.4931468136215911,
+      "grad_norm": 3.3798739910125732,
+      "learning_rate": 0.00010141302811925957,
+      "loss": 2.1009,
+      "step": 20940
+    },
+    {
+      "epoch": 0.49338231830813434,
+      "grad_norm": 2.460247039794922,
+      "learning_rate": 0.00010136592718195093,
+      "loss": 2.2822,
+      "step": 20950
+    },
+    {
+      "epoch": 0.4936178229946776,
+      "grad_norm": 2.2641265392303467,
+      "learning_rate": 0.00010131882624464228,
+      "loss": 2.0259,
+      "step": 20960
+    },
+    {
+      "epoch": 0.49385332768122087,
+      "grad_norm": 2.0152974128723145,
+      "learning_rate": 0.00010127172530733362,
+      "loss": 2.2931,
+      "step": 20970
+    },
+    {
+      "epoch": 0.49408883236776413,
+      "grad_norm": 2.167557716369629,
+      "learning_rate": 0.00010122462437002498,
+      "loss": 1.9011,
+      "step": 20980
+    },
+    {
+      "epoch": 0.4943243370543074,
+      "grad_norm": 2.114088773727417,
+      "learning_rate": 0.00010117752343271632,
+      "loss": 1.8406,
+      "step": 20990
+    },
+    {
+      "epoch": 0.49455984174085066,
+      "grad_norm": 2.915698289871216,
+      "learning_rate": 0.00010113042249540767,
+      "loss": 2.1208,
+      "step": 21000
+    },
+    {
+      "epoch": 0.4947953464273939,
+      "grad_norm": 2.237755060195923,
+      "learning_rate": 0.000101083321558099,
+      "loss": 2.0806,
+      "step": 21010
+    },
+    {
+      "epoch": 0.4950308511139372,
+      "grad_norm": 2.4972400665283203,
+      "learning_rate": 0.00010103622062079037,
+      "loss": 2.1965,
+      "step": 21020
+    },
+    {
+      "epoch": 0.49526635580048045,
+      "grad_norm": 1.718467116355896,
+      "learning_rate": 0.0001009891196834817,
+      "loss": 1.7707,
+      "step": 21030
+    },
+    {
+      "epoch": 0.4955018604870237,
+      "grad_norm": 2.2876641750335693,
+      "learning_rate": 0.00010094201874617307,
+      "loss": 2.0414,
+      "step": 21040
+    },
+    {
+      "epoch": 0.495737365173567,
+      "grad_norm": 2.5221569538116455,
+      "learning_rate": 0.0001008949178088644,
+      "loss": 1.9567,
+      "step": 21050
+    },
+    {
+      "epoch": 0.49597286986011024,
+      "grad_norm": 2.7397263050079346,
+      "learning_rate": 0.00010084781687155575,
+      "loss": 1.938,
+      "step": 21060
+    },
+    {
+      "epoch": 0.49620837454665345,
+      "grad_norm": 1.9480063915252686,
+      "learning_rate": 0.00010080071593424709,
+      "loss": 2.033,
+      "step": 21070
+    },
+    {
+      "epoch": 0.4964438792331967,
+      "grad_norm": 2.3631954193115234,
+      "learning_rate": 0.00010075361499693846,
+      "loss": 2.1295,
+      "step": 21080
+    },
+    {
+      "epoch": 0.49667938391974,
+      "grad_norm": 2.4683914184570312,
+      "learning_rate": 0.00010070651405962979,
+      "loss": 1.9156,
+      "step": 21090
+    },
+    {
+      "epoch": 0.49691488860628324,
+      "grad_norm": 2.1928350925445557,
+      "learning_rate": 0.00010065941312232114,
+      "loss": 2.1845,
+      "step": 21100
+    },
+    {
+      "epoch": 0.4971503932928265,
+      "grad_norm": 1.8933937549591064,
+      "learning_rate": 0.00010061231218501248,
+      "loss": 2.0413,
+      "step": 21110
+    },
+    {
+      "epoch": 0.4973858979793698,
+      "grad_norm": 2.532449722290039,
+      "learning_rate": 0.00010056521124770384,
+      "loss": 2.1513,
+      "step": 21120
+    },
+    {
+      "epoch": 0.49762140266591304,
+      "grad_norm": 2.1574230194091797,
+      "learning_rate": 0.00010051811031039518,
+      "loss": 1.8182,
+      "step": 21130
+    },
+    {
+      "epoch": 0.4978569073524563,
+      "grad_norm": 2.202259063720703,
+      "learning_rate": 0.00010047100937308653,
+      "loss": 2.2352,
+      "step": 21140
+    },
+    {
+      "epoch": 0.49809241203899957,
+      "grad_norm": 2.3444535732269287,
+      "learning_rate": 0.00010042390843577787,
+      "loss": 2.0119,
+      "step": 21150
+    },
+    {
+      "epoch": 0.49832791672554283,
+      "grad_norm": 2.5278213024139404,
+      "learning_rate": 0.00010037680749846923,
+      "loss": 2.1794,
+      "step": 21160
+    },
+    {
+      "epoch": 0.4985634214120861,
+      "grad_norm": 1.6672074794769287,
+      "learning_rate": 0.00010032970656116057,
+      "loss": 1.8915,
+      "step": 21170
+    },
+    {
+      "epoch": 0.49879892609862936,
+      "grad_norm": 1.9059944152832031,
+      "learning_rate": 0.00010028260562385192,
+      "loss": 1.9642,
+      "step": 21180
+    },
+    {
+      "epoch": 0.4990344307851726,
+      "grad_norm": 2.001873254776001,
+      "learning_rate": 0.00010023550468654325,
+      "loss": 2.1203,
+      "step": 21190
+    },
+    {
+      "epoch": 0.4992699354717159,
+      "grad_norm": 2.2856180667877197,
+      "learning_rate": 0.00010018840374923462,
+      "loss": 2.1251,
+      "step": 21200
+    },
+    {
+      "epoch": 0.49950544015825915,
+      "grad_norm": 3.174320936203003,
+      "learning_rate": 0.00010014130281192595,
+      "loss": 2.0888,
+      "step": 21210
+    },
+    {
+      "epoch": 0.4997409448448024,
+      "grad_norm": 3.9642059803009033,
+      "learning_rate": 0.0001000942018746173,
+      "loss": 2.13,
+      "step": 21220
+    },
+    {
+      "epoch": 0.4999764495313457,
+      "grad_norm": 2.8816041946411133,
+      "learning_rate": 0.00010004710093730864,
+      "loss": 2.2852,
+      "step": 21230
+    },
+    {
+      "epoch": 0.5002119542178889,
+      "grad_norm": 2.1514241695404053,
+      "learning_rate": 0.0001,
+      "loss": 2.175,
+      "step": 21240
+    },
+    {
+      "epoch": 0.5004474589044322,
+      "grad_norm": 2.823493242263794,
+      "learning_rate": 9.995289906269136e-05,
+      "loss": 1.9793,
+      "step": 21250
+    },
+    {
+      "epoch": 0.5006829635909754,
+      "grad_norm": 2.3700454235076904,
+      "learning_rate": 9.99057981253827e-05,
+      "loss": 2.2788,
+      "step": 21260
+    },
+    {
+      "epoch": 0.5009184682775187,
+      "grad_norm": 2.2594282627105713,
+      "learning_rate": 9.985869718807404e-05,
+      "loss": 2.0639,
+      "step": 21270
+    },
+    {
+      "epoch": 0.501153972964062,
+      "grad_norm": 2.6550183296203613,
+      "learning_rate": 9.981159625076539e-05,
+      "loss": 2.1829,
+      "step": 21280
+    },
+    {
+      "epoch": 0.5013894776506053,
+      "grad_norm": 2.1802735328674316,
+      "learning_rate": 9.976449531345674e-05,
+      "loss": 2.0109,
+      "step": 21290
+    },
+    {
+      "epoch": 0.5016249823371485,
+      "grad_norm": 2.058255672454834,
+      "learning_rate": 9.971739437614809e-05,
+      "loss": 2.2871,
+      "step": 21300
+    },
+    {
+      "epoch": 0.5018604870236918,
+      "grad_norm": 2.2070486545562744,
+      "learning_rate": 9.967029343883943e-05,
+      "loss": 2.062,
+      "step": 21310
+    },
+    {
+      "epoch": 0.502095991710235,
+      "grad_norm": 2.2400360107421875,
+      "learning_rate": 9.962319250153078e-05,
+      "loss": 2.0432,
+      "step": 21320
+    },
+    {
+      "epoch": 0.5023314963967783,
+      "grad_norm": 1.9597783088684082,
+      "learning_rate": 9.957609156422213e-05,
+      "loss": 2.0918,
+      "step": 21330
+    },
+    {
+      "epoch": 0.5025670010833215,
+      "grad_norm": 2.0394837856292725,
+      "learning_rate": 9.952899062691348e-05,
+      "loss": 2.0768,
+      "step": 21340
+    },
+    {
+      "epoch": 0.5028025057698648,
+      "grad_norm": 2.9763147830963135,
+      "learning_rate": 9.948188968960482e-05,
+      "loss": 2.2218,
+      "step": 21350
+    },
+    {
+      "epoch": 0.5030380104564081,
+      "grad_norm": 2.2641072273254395,
+      "learning_rate": 9.943478875229617e-05,
+      "loss": 2.1642,
+      "step": 21360
+    },
+    {
+      "epoch": 0.5032735151429514,
+      "grad_norm": 1.8236515522003174,
+      "learning_rate": 9.938768781498752e-05,
+      "loss": 2.0839,
+      "step": 21370
+    },
+    {
+      "epoch": 0.5035090198294946,
+      "grad_norm": 2.490896224975586,
+      "learning_rate": 9.934058687767887e-05,
+      "loss": 2.2359,
+      "step": 21380
+    },
+    {
+      "epoch": 0.5037445245160379,
+      "grad_norm": 2.3176302909851074,
+      "learning_rate": 9.929348594037022e-05,
+      "loss": 2.2307,
+      "step": 21390
+    },
+    {
+      "epoch": 0.5039800292025811,
+      "grad_norm": 2.2265188694000244,
+      "learning_rate": 9.924638500306157e-05,
+      "loss": 2.0561,
+      "step": 21400
+    },
+    {
+      "epoch": 0.5042155338891244,
+      "grad_norm": 2.182687759399414,
+      "learning_rate": 9.919928406575292e-05,
+      "loss": 2.1138,
+      "step": 21410
+    },
+    {
+      "epoch": 0.5044510385756676,
+      "grad_norm": 2.3917243480682373,
+      "learning_rate": 9.915218312844427e-05,
+      "loss": 2.1989,
+      "step": 21420
+    },
+    {
+      "epoch": 0.504686543262211,
+      "grad_norm": 2.182008981704712,
+      "learning_rate": 9.91050821911356e-05,
+      "loss": 1.9135,
+      "step": 21430
+    },
+    {
+      "epoch": 0.5049220479487542,
+      "grad_norm": 2.1667606830596924,
+      "learning_rate": 9.905798125382696e-05,
+      "loss": 1.977,
+      "step": 21440
+    },
+    {
+      "epoch": 0.5051575526352975,
+      "grad_norm": 2.5774123668670654,
+      "learning_rate": 9.90108803165183e-05,
+      "loss": 2.0104,
+      "step": 21450
+    },
+    {
+      "epoch": 0.5053930573218407,
+      "grad_norm": 2.3878164291381836,
+      "learning_rate": 9.896377937920966e-05,
+      "loss": 1.9225,
+      "step": 21460
+    },
+    {
+      "epoch": 0.505628562008384,
+      "grad_norm": 2.676603078842163,
+      "learning_rate": 9.8916678441901e-05,
+      "loss": 2.1486,
+      "step": 21470
+    },
+    {
+      "epoch": 0.5058640666949272,
+      "grad_norm": 1.9781603813171387,
+      "learning_rate": 9.886957750459234e-05,
+      "loss": 2.0173,
+      "step": 21480
+    },
+    {
+      "epoch": 0.5060995713814704,
+      "grad_norm": 2.2742934226989746,
+      "learning_rate": 9.882247656728369e-05,
+      "loss": 2.0306,
+      "step": 21490
+    },
+    {
+      "epoch": 0.5063350760680138,
+      "grad_norm": 1.8691270351409912,
+      "learning_rate": 9.877537562997504e-05,
+      "loss": 1.8714,
+      "step": 21500
+    },
+    {
+      "epoch": 0.506570580754557,
+      "grad_norm": 2.0995771884918213,
+      "learning_rate": 9.87282746926664e-05,
+      "loss": 1.931,
+      "step": 21510
+    },
+    {
+      "epoch": 0.5068060854411003,
+      "grad_norm": 2.2483112812042236,
+      "learning_rate": 9.868117375535773e-05,
+      "loss": 2.1824,
+      "step": 21520
+    },
+    {
+      "epoch": 0.5070415901276435,
+      "grad_norm": 2.2055604457855225,
+      "learning_rate": 9.863407281804908e-05,
+      "loss": 1.8965,
+      "step": 21530
+    },
+    {
+      "epoch": 0.5072770948141868,
+      "grad_norm": 2.187610626220703,
+      "learning_rate": 9.858697188074043e-05,
+      "loss": 2.1238,
+      "step": 21540
+    },
+    {
+      "epoch": 0.50751259950073,
+      "grad_norm": 2.0155301094055176,
+      "learning_rate": 9.853987094343178e-05,
+      "loss": 2.2547,
+      "step": 21550
+    },
+    {
+      "epoch": 0.5077481041872733,
+      "grad_norm": 2.0881471633911133,
+      "learning_rate": 9.849277000612313e-05,
+      "loss": 2.05,
+      "step": 21560
+    },
+    {
+      "epoch": 0.5079836088738166,
+      "grad_norm": 2.319646120071411,
+      "learning_rate": 9.844566906881447e-05,
+      "loss": 1.9783,
+      "step": 21570
+    },
+    {
+      "epoch": 0.5082191135603599,
+      "grad_norm": 2.136197805404663,
+      "learning_rate": 9.839856813150582e-05,
+      "loss": 2.0,
+      "step": 21580
+    },
+    {
+      "epoch": 0.5084546182469031,
+      "grad_norm": 2.998568058013916,
+      "learning_rate": 9.835146719419717e-05,
+      "loss": 2.1183,
+      "step": 21590
+    },
+    {
+      "epoch": 0.5086901229334464,
+      "grad_norm": 2.355353355407715,
+      "learning_rate": 9.830436625688852e-05,
+      "loss": 2.2414,
+      "step": 21600
+    },
+    {
+      "epoch": 0.5089256276199896,
+      "grad_norm": 1.928250789642334,
+      "learning_rate": 9.825726531957986e-05,
+      "loss": 1.8088,
+      "step": 21610
+    },
+    {
+      "epoch": 0.5091611323065329,
+      "grad_norm": 3.0732262134552,
+      "learning_rate": 9.82101643822712e-05,
+      "loss": 1.9908,
+      "step": 21620
+    },
+    {
+      "epoch": 0.5093966369930761,
+      "grad_norm": 3.366936445236206,
+      "learning_rate": 9.816306344496256e-05,
+      "loss": 2.067,
+      "step": 21630
+    },
+    {
+      "epoch": 0.5096321416796195,
+      "grad_norm": 2.2484872341156006,
+      "learning_rate": 9.81159625076539e-05,
+      "loss": 2.1143,
+      "step": 21640
+    },
+    {
+      "epoch": 0.5098676463661627,
+      "grad_norm": 2.183173894882202,
+      "learning_rate": 9.806886157034524e-05,
+      "loss": 2.1001,
+      "step": 21650
+    },
+    {
+      "epoch": 0.510103151052706,
+      "grad_norm": 2.780048131942749,
+      "learning_rate": 9.802176063303659e-05,
+      "loss": 2.2141,
+      "step": 21660
+    },
+    {
+      "epoch": 0.5103386557392492,
+      "grad_norm": 2.5140750408172607,
+      "learning_rate": 9.797465969572796e-05,
+      "loss": 2.136,
+      "step": 21670
+    },
+    {
+      "epoch": 0.5105741604257925,
+      "grad_norm": 2.4135138988494873,
+      "learning_rate": 9.792755875841931e-05,
+      "loss": 1.924,
+      "step": 21680
+    },
+    {
+      "epoch": 0.5108096651123357,
+      "grad_norm": 2.8856024742126465,
+      "learning_rate": 9.788045782111064e-05,
+      "loss": 1.9122,
+      "step": 21690
+    },
+    {
+      "epoch": 0.511045169798879,
+      "grad_norm": 2.015821695327759,
+      "learning_rate": 9.7833356883802e-05,
+      "loss": 2.2464,
+      "step": 21700
+    },
+    {
+      "epoch": 0.5112806744854222,
+      "grad_norm": 3.2356677055358887,
+      "learning_rate": 9.778625594649334e-05,
+      "loss": 2.0949,
+      "step": 21710
+    },
+    {
+      "epoch": 0.5115161791719656,
+      "grad_norm": 1.8908910751342773,
+      "learning_rate": 9.77391550091847e-05,
+      "loss": 1.8944,
+      "step": 21720
+    },
+    {
+      "epoch": 0.5117516838585088,
+      "grad_norm": 1.9458872079849243,
+      "learning_rate": 9.769205407187603e-05,
+      "loss": 2.1298,
+      "step": 21730
+    },
+    {
+      "epoch": 0.5119871885450521,
+      "grad_norm": 2.1977505683898926,
+      "learning_rate": 9.764495313456738e-05,
+      "loss": 2.3225,
+      "step": 21740
+    },
+    {
+      "epoch": 0.5122226932315953,
+      "grad_norm": 2.247251033782959,
+      "learning_rate": 9.759785219725873e-05,
+      "loss": 2.0982,
+      "step": 21750
+    },
+    {
+      "epoch": 0.5124581979181386,
+      "grad_norm": 2.334291934967041,
+      "learning_rate": 9.755075125995008e-05,
+      "loss": 2.144,
+      "step": 21760
+    },
+    {
+      "epoch": 0.5126937026046818,
+      "grad_norm": 2.0786211490631104,
+      "learning_rate": 9.750365032264143e-05,
+      "loss": 2.1501,
+      "step": 21770
+    },
+    {
+      "epoch": 0.512929207291225,
+      "grad_norm": 2.1926138401031494,
+      "learning_rate": 9.745654938533277e-05,
+      "loss": 2.0627,
+      "step": 21780
+    },
+    {
+      "epoch": 0.5131647119777684,
+      "grad_norm": 2.517853021621704,
+      "learning_rate": 9.740944844802412e-05,
+      "loss": 2.0631,
+      "step": 21790
+    },
+    {
+      "epoch": 0.5134002166643116,
+      "grad_norm": 1.8308531045913696,
+      "learning_rate": 9.736234751071547e-05,
+      "loss": 2.1762,
+      "step": 21800
+    },
+    {
+      "epoch": 0.5136357213508549,
+      "grad_norm": 2.1084744930267334,
+      "learning_rate": 9.731524657340682e-05,
+      "loss": 1.9098,
+      "step": 21810
+    },
+    {
+      "epoch": 0.5138712260373981,
+      "grad_norm": 2.3798017501831055,
+      "learning_rate": 9.726814563609816e-05,
+      "loss": 1.8055,
+      "step": 21820
+    },
+    {
+      "epoch": 0.5141067307239414,
+      "grad_norm": 2.443279266357422,
+      "learning_rate": 9.72210446987895e-05,
+      "loss": 2.0454,
+      "step": 21830
+    },
+    {
+      "epoch": 0.5143422354104846,
+      "grad_norm": 2.0576086044311523,
+      "learning_rate": 9.717394376148086e-05,
+      "loss": 2.2606,
+      "step": 21840
+    },
+    {
+      "epoch": 0.514577740097028,
+      "grad_norm": 1.869069218635559,
+      "learning_rate": 9.712684282417221e-05,
+      "loss": 1.8852,
+      "step": 21850
+    },
+    {
+      "epoch": 0.5148132447835712,
+      "grad_norm": 1.9165115356445312,
+      "learning_rate": 9.707974188686356e-05,
+      "loss": 1.9244,
+      "step": 21860
+    },
+    {
+      "epoch": 0.5150487494701145,
+      "grad_norm": 1.7431786060333252,
+      "learning_rate": 9.70326409495549e-05,
+      "loss": 1.9097,
+      "step": 21870
+    },
+    {
+      "epoch": 0.5152842541566577,
+      "grad_norm": 2.2474775314331055,
+      "learning_rate": 9.698554001224624e-05,
+      "loss": 2.1901,
+      "step": 21880
+    },
+    {
+      "epoch": 0.515519758843201,
+      "grad_norm": 1.9225223064422607,
+      "learning_rate": 9.69384390749376e-05,
+      "loss": 2.217,
+      "step": 21890
+    },
+    {
+      "epoch": 0.5157552635297442,
+      "grad_norm": 1.9866782426834106,
+      "learning_rate": 9.689133813762894e-05,
+      "loss": 1.8987,
+      "step": 21900
+    },
+    {
+      "epoch": 0.5159907682162875,
+      "grad_norm": 2.126774549484253,
+      "learning_rate": 9.684423720032028e-05,
+      "loss": 2.0953,
+      "step": 21910
+    },
+    {
+      "epoch": 0.5162262729028307,
+      "grad_norm": 1.8921127319335938,
+      "learning_rate": 9.679713626301163e-05,
+      "loss": 2.0589,
+      "step": 21920
+    },
+    {
+      "epoch": 0.5164617775893741,
+      "grad_norm": 2.78843092918396,
+      "learning_rate": 9.675003532570298e-05,
+      "loss": 1.9555,
+      "step": 21930
+    },
+    {
+      "epoch": 0.5166972822759173,
+      "grad_norm": 2.5705997943878174,
+      "learning_rate": 9.670293438839433e-05,
+      "loss": 2.0471,
+      "step": 21940
+    },
+    {
+      "epoch": 0.5169327869624606,
+      "grad_norm": 1.8780876398086548,
+      "learning_rate": 9.665583345108568e-05,
+      "loss": 2.0409,
+      "step": 21950
+    },
+    {
+      "epoch": 0.5171682916490038,
+      "grad_norm": 2.1440744400024414,
+      "learning_rate": 9.660873251377703e-05,
+      "loss": 2.0148,
+      "step": 21960
+    },
+    {
+      "epoch": 0.5174037963355471,
+      "grad_norm": 2.174579381942749,
+      "learning_rate": 9.656163157646838e-05,
+      "loss": 2.1547,
+      "step": 21970
+    },
+    {
+      "epoch": 0.5176393010220903,
+      "grad_norm": 3.240619421005249,
+      "learning_rate": 9.651453063915973e-05,
+      "loss": 2.0667,
+      "step": 21980
+    },
+    {
+      "epoch": 0.5178748057086336,
+      "grad_norm": 2.3205692768096924,
+      "learning_rate": 9.646742970185107e-05,
+      "loss": 1.8602,
+      "step": 21990
+    },
+    {
+      "epoch": 0.5181103103951769,
+      "grad_norm": 3.049022912979126,
+      "learning_rate": 9.642032876454242e-05,
+      "loss": 2.0942,
+      "step": 22000
+    },
+    {
+      "epoch": 0.5183458150817202,
+      "grad_norm": 2.585932970046997,
+      "learning_rate": 9.637322782723377e-05,
+      "loss": 1.9998,
+      "step": 22010
+    },
+    {
+      "epoch": 0.5185813197682634,
+      "grad_norm": 2.4109232425689697,
+      "learning_rate": 9.632612688992512e-05,
+      "loss": 2.075,
+      "step": 22020
+    },
+    {
+      "epoch": 0.5188168244548067,
+      "grad_norm": 2.334937334060669,
+      "learning_rate": 9.627902595261646e-05,
+      "loss": 1.8412,
+      "step": 22030
+    },
+    {
+      "epoch": 0.5190523291413499,
+      "grad_norm": 3.0095884799957275,
+      "learning_rate": 9.623192501530781e-05,
+      "loss": 1.918,
+      "step": 22040
+    },
+    {
+      "epoch": 0.5192878338278932,
+      "grad_norm": 1.9258203506469727,
+      "learning_rate": 9.618482407799916e-05,
+      "loss": 1.9809,
+      "step": 22050
+    },
+    {
+      "epoch": 0.5195233385144364,
+      "grad_norm": 1.8226529359817505,
+      "learning_rate": 9.613772314069051e-05,
+      "loss": 1.9827,
+      "step": 22060
+    },
+    {
+      "epoch": 0.5197588432009796,
+      "grad_norm": 2.051090955734253,
+      "learning_rate": 9.609062220338186e-05,
+      "loss": 1.9504,
+      "step": 22070
+    },
+    {
+      "epoch": 0.519994347887523,
+      "grad_norm": 2.140054941177368,
+      "learning_rate": 9.60435212660732e-05,
+      "loss": 2.0869,
+      "step": 22080
+    },
+    {
+      "epoch": 0.5202298525740662,
+      "grad_norm": 2.1044740676879883,
+      "learning_rate": 9.599642032876455e-05,
+      "loss": 1.9475,
+      "step": 22090
+    },
+    {
+      "epoch": 0.5204653572606095,
+      "grad_norm": 1.9756555557250977,
+      "learning_rate": 9.59493193914559e-05,
+      "loss": 2.0775,
+      "step": 22100
+    },
+    {
+      "epoch": 0.5207008619471527,
+      "grad_norm": 2.549635171890259,
+      "learning_rate": 9.590221845414725e-05,
+      "loss": 1.8268,
+      "step": 22110
+    },
+    {
+      "epoch": 0.520936366633696,
+      "grad_norm": 2.121861696243286,
+      "learning_rate": 9.585511751683858e-05,
+      "loss": 2.103,
+      "step": 22120
+    },
+    {
+      "epoch": 0.5211718713202392,
+      "grad_norm": 2.389665365219116,
+      "learning_rate": 9.580801657952993e-05,
+      "loss": 2.2068,
+      "step": 22130
+    },
+    {
+      "epoch": 0.5214073760067826,
+      "grad_norm": 2.340815782546997,
+      "learning_rate": 9.576091564222128e-05,
+      "loss": 2.0227,
+      "step": 22140
+    },
+    {
+      "epoch": 0.5216428806933258,
+      "grad_norm": 2.191857099533081,
+      "learning_rate": 9.571381470491263e-05,
+      "loss": 2.1419,
+      "step": 22150
+    },
+    {
+      "epoch": 0.5218783853798691,
+      "grad_norm": 3.036268711090088,
+      "learning_rate": 9.566671376760398e-05,
+      "loss": 2.0012,
+      "step": 22160
+    },
+    {
+      "epoch": 0.5221138900664123,
+      "grad_norm": 2.6708240509033203,
+      "learning_rate": 9.561961283029532e-05,
+      "loss": 2.052,
+      "step": 22170
+    },
+    {
+      "epoch": 0.5223493947529556,
+      "grad_norm": 2.439678907394409,
+      "learning_rate": 9.557251189298667e-05,
+      "loss": 2.2863,
+      "step": 22180
+    },
+    {
+      "epoch": 0.5225848994394988,
+      "grad_norm": 2.5144474506378174,
+      "learning_rate": 9.552541095567802e-05,
+      "loss": 2.1124,
+      "step": 22190
+    },
+    {
+      "epoch": 0.5228204041260421,
+      "grad_norm": 2.6806702613830566,
+      "learning_rate": 9.547831001836937e-05,
+      "loss": 2.1127,
+      "step": 22200
+    },
+    {
+      "epoch": 0.5230559088125853,
+      "grad_norm": 2.2948286533355713,
+      "learning_rate": 9.543120908106071e-05,
+      "loss": 2.0362,
+      "step": 22210
+    },
+    {
+      "epoch": 0.5232914134991287,
+      "grad_norm": 2.144993543624878,
+      "learning_rate": 9.538410814375206e-05,
+      "loss": 1.8412,
+      "step": 22220
+    },
+    {
+      "epoch": 0.5235269181856719,
+      "grad_norm": 2.0969650745391846,
+      "learning_rate": 9.533700720644342e-05,
+      "loss": 2.1297,
+      "step": 22230
+    },
+    {
+      "epoch": 0.5237624228722152,
+      "grad_norm": 1.9239368438720703,
+      "learning_rate": 9.528990626913477e-05,
+      "loss": 2.0009,
+      "step": 22240
+    },
+    {
+      "epoch": 0.5239979275587584,
+      "grad_norm": 2.1357004642486572,
+      "learning_rate": 9.524280533182611e-05,
+      "loss": 2.1151,
+      "step": 22250
+    },
+    {
+      "epoch": 0.5242334322453017,
+      "grad_norm": 2.1568477153778076,
+      "learning_rate": 9.519570439451746e-05,
+      "loss": 1.991,
+      "step": 22260
+    },
+    {
+      "epoch": 0.5244689369318449,
+      "grad_norm": 2.371934175491333,
+      "learning_rate": 9.514860345720881e-05,
+      "loss": 2.1708,
+      "step": 22270
+    },
+    {
+      "epoch": 0.5247044416183883,
+      "grad_norm": 1.8997218608856201,
+      "learning_rate": 9.510150251990016e-05,
+      "loss": 2.0686,
+      "step": 22280
+    },
+    {
+      "epoch": 0.5249399463049315,
+      "grad_norm": 2.1850790977478027,
+      "learning_rate": 9.50544015825915e-05,
+      "loss": 2.0206,
+      "step": 22290
+    },
+    {
+      "epoch": 0.5251754509914748,
+      "grad_norm": 2.4998154640197754,
+      "learning_rate": 9.500730064528285e-05,
+      "loss": 1.874,
+      "step": 22300
+    },
+    {
+      "epoch": 0.525410955678018,
+      "grad_norm": 2.6623775959014893,
+      "learning_rate": 9.49601997079742e-05,
+      "loss": 2.1144,
+      "step": 22310
+    },
+    {
+      "epoch": 0.5256464603645613,
+      "grad_norm": 2.4677085876464844,
+      "learning_rate": 9.491309877066555e-05,
+      "loss": 2.0148,
+      "step": 22320
+    },
+    {
+      "epoch": 0.5258819650511045,
+      "grad_norm": 2.5146234035491943,
+      "learning_rate": 9.486599783335688e-05,
+      "loss": 2.0617,
+      "step": 22330
+    },
+    {
+      "epoch": 0.5261174697376477,
+      "grad_norm": 1.9805279970169067,
+      "learning_rate": 9.481889689604823e-05,
+      "loss": 2.0317,
+      "step": 22340
+    },
+    {
+      "epoch": 0.526352974424191,
+      "grad_norm": 1.9576618671417236,
+      "learning_rate": 9.477179595873958e-05,
+      "loss": 2.1339,
+      "step": 22350
+    },
+    {
+      "epoch": 0.5265884791107343,
+      "grad_norm": 2.2992379665374756,
+      "learning_rate": 9.472469502143093e-05,
+      "loss": 2.1361,
+      "step": 22360
+    },
+    {
+      "epoch": 0.5268239837972776,
+      "grad_norm": 2.034731149673462,
+      "learning_rate": 9.467759408412228e-05,
+      "loss": 2.062,
+      "step": 22370
+    },
+    {
+      "epoch": 0.5270594884838208,
+      "grad_norm": 2.1001486778259277,
+      "learning_rate": 9.463049314681362e-05,
+      "loss": 2.1102,
+      "step": 22380
+    },
+    {
+      "epoch": 0.5272949931703641,
+      "grad_norm": 1.7902547121047974,
+      "learning_rate": 9.458339220950497e-05,
+      "loss": 1.8913,
+      "step": 22390
+    },
+    {
+      "epoch": 0.5275304978569073,
+      "grad_norm": 2.0695128440856934,
+      "learning_rate": 9.453629127219632e-05,
+      "loss": 2.0724,
+      "step": 22400
+    },
+    {
+      "epoch": 0.5277660025434506,
+      "grad_norm": 2.392047643661499,
+      "learning_rate": 9.448919033488767e-05,
+      "loss": 1.8279,
+      "step": 22410
+    },
+    {
+      "epoch": 0.5280015072299938,
+      "grad_norm": 2.0684807300567627,
+      "learning_rate": 9.444208939757901e-05,
+      "loss": 2.2671,
+      "step": 22420
+    },
+    {
+      "epoch": 0.5282370119165372,
+      "grad_norm": 1.8215241432189941,
+      "learning_rate": 9.439498846027036e-05,
+      "loss": 2.1644,
+      "step": 22430
+    },
+    {
+      "epoch": 0.5284725166030804,
+      "grad_norm": 2.322417974472046,
+      "learning_rate": 9.434788752296171e-05,
+      "loss": 2.1318,
+      "step": 22440
+    },
+    {
+      "epoch": 0.5287080212896237,
+      "grad_norm": 2.333282232284546,
+      "learning_rate": 9.430078658565306e-05,
+      "loss": 1.8483,
+      "step": 22450
+    },
+    {
+      "epoch": 0.5289435259761669,
+      "grad_norm": 2.832303285598755,
+      "learning_rate": 9.425368564834441e-05,
+      "loss": 2.0394,
+      "step": 22460
+    },
+    {
+      "epoch": 0.5291790306627102,
+      "grad_norm": 1.6820281744003296,
+      "learning_rate": 9.420658471103575e-05,
+      "loss": 2.027,
+      "step": 22470
+    },
+    {
+      "epoch": 0.5294145353492534,
+      "grad_norm": 3.3374030590057373,
+      "learning_rate": 9.41594837737271e-05,
+      "loss": 2.1277,
+      "step": 22480
+    },
+    {
+      "epoch": 0.5296500400357967,
+      "grad_norm": 2.0465030670166016,
+      "learning_rate": 9.411238283641845e-05,
+      "loss": 2.0185,
+      "step": 22490
+    },
+    {
+      "epoch": 0.52988554472234,
+      "grad_norm": 2.40169620513916,
+      "learning_rate": 9.40652818991098e-05,
+      "loss": 2.2946,
+      "step": 22500
+    },
+    {
+      "epoch": 0.5301210494088833,
+      "grad_norm": 2.6130552291870117,
+      "learning_rate": 9.401818096180113e-05,
+      "loss": 2.3308,
+      "step": 22510
+    },
+    {
+      "epoch": 0.5303565540954265,
+      "grad_norm": 2.0632524490356445,
+      "learning_rate": 9.39710800244925e-05,
+      "loss": 2.1752,
+      "step": 22520
+    },
+    {
+      "epoch": 0.5305920587819698,
+      "grad_norm": 1.7699297666549683,
+      "learning_rate": 9.392397908718385e-05,
+      "loss": 2.0812,
+      "step": 22530
+    },
+    {
+      "epoch": 0.530827563468513,
+      "grad_norm": 2.3439061641693115,
+      "learning_rate": 9.38768781498752e-05,
+      "loss": 2.0295,
+      "step": 22540
+    },
+    {
+      "epoch": 0.5310630681550563,
+      "grad_norm": 2.736743688583374,
+      "learning_rate": 9.382977721256653e-05,
+      "loss": 2.0404,
+      "step": 22550
+    },
+    {
+      "epoch": 0.5312985728415995,
+      "grad_norm": 1.7512779235839844,
+      "learning_rate": 9.378267627525788e-05,
+      "loss": 2.0293,
+      "step": 22560
+    },
+    {
+      "epoch": 0.5315340775281429,
+      "grad_norm": 2.9181957244873047,
+      "learning_rate": 9.373557533794923e-05,
+      "loss": 2.0772,
+      "step": 22570
+    },
+    {
+      "epoch": 0.5317695822146861,
+      "grad_norm": 2.2531585693359375,
+      "learning_rate": 9.368847440064059e-05,
+      "loss": 2.1138,
+      "step": 22580
+    },
+    {
+      "epoch": 0.5320050869012294,
+      "grad_norm": 2.3378305435180664,
+      "learning_rate": 9.364137346333192e-05,
+      "loss": 2.164,
+      "step": 22590
+    },
+    {
+      "epoch": 0.5322405915877726,
+      "grad_norm": 2.257312536239624,
+      "learning_rate": 9.359427252602327e-05,
+      "loss": 1.9478,
+      "step": 22600
+    },
+    {
+      "epoch": 0.5324760962743159,
+      "grad_norm": 3.4792850017547607,
+      "learning_rate": 9.354717158871462e-05,
+      "loss": 1.9697,
+      "step": 22610
+    },
+    {
+      "epoch": 0.5327116009608591,
+      "grad_norm": 2.011868476867676,
+      "learning_rate": 9.350007065140597e-05,
+      "loss": 1.9607,
+      "step": 22620
+    },
+    {
+      "epoch": 0.5329471056474023,
+      "grad_norm": 3.508305072784424,
+      "learning_rate": 9.345296971409731e-05,
+      "loss": 2.0881,
+      "step": 22630
+    },
+    {
+      "epoch": 0.5331826103339457,
+      "grad_norm": 2.206843137741089,
+      "learning_rate": 9.340586877678866e-05,
+      "loss": 2.0199,
+      "step": 22640
+    },
+    {
+      "epoch": 0.5334181150204889,
+      "grad_norm": 2.1505041122436523,
+      "learning_rate": 9.335876783948001e-05,
+      "loss": 1.9688,
+      "step": 22650
+    },
+    {
+      "epoch": 0.5336536197070322,
+      "grad_norm": 2.0961718559265137,
+      "learning_rate": 9.331166690217136e-05,
+      "loss": 2.0561,
+      "step": 22660
+    },
+    {
+      "epoch": 0.5338891243935754,
+      "grad_norm": 1.9985352754592896,
+      "learning_rate": 9.326456596486271e-05,
+      "loss": 1.9848,
+      "step": 22670
+    },
+    {
+      "epoch": 0.5341246290801187,
+      "grad_norm": 3.043963670730591,
+      "learning_rate": 9.321746502755405e-05,
+      "loss": 2.0881,
+      "step": 22680
+    },
+    {
+      "epoch": 0.5343601337666619,
+      "grad_norm": 2.3572945594787598,
+      "learning_rate": 9.31703640902454e-05,
+      "loss": 1.995,
+      "step": 22690
+    },
+    {
+      "epoch": 0.5345956384532052,
+      "grad_norm": 2.074956178665161,
+      "learning_rate": 9.312326315293675e-05,
+      "loss": 2.0339,
+      "step": 22700
+    },
+    {
+      "epoch": 0.5348311431397484,
+      "grad_norm": 1.7180410623550415,
+      "learning_rate": 9.30761622156281e-05,
+      "loss": 1.9256,
+      "step": 22710
+    },
+    {
+      "epoch": 0.5350666478262918,
+      "grad_norm": 2.5363078117370605,
+      "learning_rate": 9.302906127831943e-05,
+      "loss": 2.2295,
+      "step": 22720
+    },
+    {
+      "epoch": 0.535302152512835,
+      "grad_norm": 2.200470209121704,
+      "learning_rate": 9.298196034101078e-05,
+      "loss": 1.904,
+      "step": 22730
+    },
+    {
+      "epoch": 0.5355376571993783,
+      "grad_norm": 1.8723726272583008,
+      "learning_rate": 9.293485940370213e-05,
+      "loss": 2.2491,
+      "step": 22740
+    },
+    {
+      "epoch": 0.5357731618859215,
+      "grad_norm": 2.703275442123413,
+      "learning_rate": 9.288775846639349e-05,
+      "loss": 1.7582,
+      "step": 22750
+    },
+    {
+      "epoch": 0.5360086665724648,
+      "grad_norm": 2.2764389514923096,
+      "learning_rate": 9.284065752908484e-05,
+      "loss": 1.9832,
+      "step": 22760
+    },
+    {
+      "epoch": 0.536244171259008,
+      "grad_norm": 2.089346170425415,
+      "learning_rate": 9.279355659177617e-05,
+      "loss": 2.2175,
+      "step": 22770
+    },
+    {
+      "epoch": 0.5364796759455513,
+      "grad_norm": 2.090502977371216,
+      "learning_rate": 9.274645565446752e-05,
+      "loss": 2.1681,
+      "step": 22780
+    },
+    {
+      "epoch": 0.5367151806320946,
+      "grad_norm": 3.0695157051086426,
+      "learning_rate": 9.269935471715887e-05,
+      "loss": 2.205,
+      "step": 22790
+    },
+    {
+      "epoch": 0.5369506853186379,
+      "grad_norm": 2.0072319507598877,
+      "learning_rate": 9.265225377985022e-05,
+      "loss": 1.9657,
+      "step": 22800
+    },
+    {
+      "epoch": 0.5371861900051811,
+      "grad_norm": 2.8740854263305664,
+      "learning_rate": 9.260515284254157e-05,
+      "loss": 2.011,
+      "step": 22810
+    },
+    {
+      "epoch": 0.5374216946917244,
+      "grad_norm": 2.9721429347991943,
+      "learning_rate": 9.255805190523292e-05,
+      "loss": 2.0737,
+      "step": 22820
+    },
+    {
+      "epoch": 0.5376571993782676,
+      "grad_norm": 1.8961126804351807,
+      "learning_rate": 9.251095096792427e-05,
+      "loss": 1.89,
+      "step": 22830
+    },
+    {
+      "epoch": 0.5378927040648109,
+      "grad_norm": 2.3074028491973877,
+      "learning_rate": 9.246385003061562e-05,
+      "loss": 2.0739,
+      "step": 22840
+    },
+    {
+      "epoch": 0.5381282087513541,
+      "grad_norm": 2.082486391067505,
+      "learning_rate": 9.241674909330696e-05,
+      "loss": 2.0767,
+      "step": 22850
+    },
+    {
+      "epoch": 0.5383637134378975,
+      "grad_norm": 2.289553165435791,
+      "learning_rate": 9.236964815599831e-05,
+      "loss": 2.1389,
+      "step": 22860
+    },
+    {
+      "epoch": 0.5385992181244407,
+      "grad_norm": 2.9601359367370605,
+      "learning_rate": 9.232254721868966e-05,
+      "loss": 1.984,
+      "step": 22870
+    },
+    {
+      "epoch": 0.538834722810984,
+      "grad_norm": 2.5509331226348877,
+      "learning_rate": 9.227544628138101e-05,
+      "loss": 2.1656,
+      "step": 22880
+    },
+    {
+      "epoch": 0.5390702274975272,
+      "grad_norm": 2.1479685306549072,
+      "learning_rate": 9.222834534407235e-05,
+      "loss": 2.0994,
+      "step": 22890
+    },
+    {
+      "epoch": 0.5393057321840704,
+      "grad_norm": 1.8056848049163818,
+      "learning_rate": 9.21812444067637e-05,
+      "loss": 1.9269,
+      "step": 22900
+    },
+    {
+      "epoch": 0.5395412368706137,
+      "grad_norm": 2.1940340995788574,
+      "learning_rate": 9.213414346945505e-05,
+      "loss": 1.9795,
+      "step": 22910
+    },
+    {
+      "epoch": 0.5397767415571569,
+      "grad_norm": 2.247976541519165,
+      "learning_rate": 9.20870425321464e-05,
+      "loss": 2.0206,
+      "step": 22920
+    },
+    {
+      "epoch": 0.5400122462437003,
+      "grad_norm": 2.516195774078369,
+      "learning_rate": 9.203994159483774e-05,
+      "loss": 1.9513,
+      "step": 22930
+    },
+    {
+      "epoch": 0.5402477509302435,
+      "grad_norm": 1.9935544729232788,
+      "learning_rate": 9.199284065752909e-05,
+      "loss": 1.8158,
+      "step": 22940
+    },
+    {
+      "epoch": 0.5404832556167868,
+      "grad_norm": 1.9285348653793335,
+      "learning_rate": 9.194573972022044e-05,
+      "loss": 2.0684,
+      "step": 22950
+    },
+    {
+      "epoch": 0.54071876030333,
+      "grad_norm": 2.0052757263183594,
+      "learning_rate": 9.189863878291179e-05,
+      "loss": 1.8693,
+      "step": 22960
+    },
+    {
+      "epoch": 0.5409542649898733,
+      "grad_norm": 3.399982213973999,
+      "learning_rate": 9.185153784560314e-05,
+      "loss": 2.2221,
+      "step": 22970
+    },
+    {
+      "epoch": 0.5411897696764165,
+      "grad_norm": 1.7858917713165283,
+      "learning_rate": 9.180443690829447e-05,
+      "loss": 1.9837,
+      "step": 22980
+    },
+    {
+      "epoch": 0.5414252743629598,
+      "grad_norm": 2.4309234619140625,
+      "learning_rate": 9.175733597098582e-05,
+      "loss": 2.0491,
+      "step": 22990
+    },
+    {
+      "epoch": 0.541660779049503,
+      "grad_norm": 1.9532793760299683,
+      "learning_rate": 9.171023503367717e-05,
+      "loss": 2.0639,
+      "step": 23000
+    },
+    {
+      "epoch": 0.5418962837360464,
+      "grad_norm": 2.7318344116210938,
+      "learning_rate": 9.166313409636852e-05,
+      "loss": 2.1516,
+      "step": 23010
+    },
+    {
+      "epoch": 0.5421317884225896,
+      "grad_norm": 1.8187711238861084,
+      "learning_rate": 9.161603315905986e-05,
+      "loss": 1.8155,
+      "step": 23020
+    },
+    {
+      "epoch": 0.5423672931091329,
+      "grad_norm": 2.425743818283081,
+      "learning_rate": 9.156893222175121e-05,
+      "loss": 2.1712,
+      "step": 23030
+    },
+    {
+      "epoch": 0.5426027977956761,
+      "grad_norm": 1.8586986064910889,
+      "learning_rate": 9.152183128444256e-05,
+      "loss": 2.03,
+      "step": 23040
+    },
+    {
+      "epoch": 0.5428383024822194,
+      "grad_norm": 2.399402618408203,
+      "learning_rate": 9.147473034713391e-05,
+      "loss": 1.9666,
+      "step": 23050
+    },
+    {
+      "epoch": 0.5430738071687626,
+      "grad_norm": 1.966139554977417,
+      "learning_rate": 9.142762940982526e-05,
+      "loss": 1.9428,
+      "step": 23060
+    },
+    {
+      "epoch": 0.543309311855306,
+      "grad_norm": 2.4574363231658936,
+      "learning_rate": 9.13805284725166e-05,
+      "loss": 2.115,
+      "step": 23070
+    },
+    {
+      "epoch": 0.5435448165418492,
+      "grad_norm": 2.235666275024414,
+      "learning_rate": 9.133342753520796e-05,
+      "loss": 2.0983,
+      "step": 23080
+    },
+    {
+      "epoch": 0.5437803212283925,
+      "grad_norm": 1.8291510343551636,
+      "learning_rate": 9.128632659789931e-05,
+      "loss": 2.0416,
+      "step": 23090
+    },
+    {
+      "epoch": 0.5440158259149357,
+      "grad_norm": 2.8294384479522705,
+      "learning_rate": 9.123922566059065e-05,
+      "loss": 2.2959,
+      "step": 23100
+    },
+    {
+      "epoch": 0.544251330601479,
+      "grad_norm": 2.171060085296631,
+      "learning_rate": 9.1192124723282e-05,
+      "loss": 2.1052,
+      "step": 23110
+    },
+    {
+      "epoch": 0.5444868352880222,
+      "grad_norm": 2.0458946228027344,
+      "learning_rate": 9.114502378597335e-05,
+      "loss": 2.029,
+      "step": 23120
+    },
+    {
+      "epoch": 0.5447223399745655,
+      "grad_norm": 1.9284874200820923,
+      "learning_rate": 9.10979228486647e-05,
+      "loss": 2.0337,
+      "step": 23130
+    },
+    {
+      "epoch": 0.5449578446611087,
+      "grad_norm": 1.6422959566116333,
+      "learning_rate": 9.105082191135605e-05,
+      "loss": 2.2123,
+      "step": 23140
+    },
+    {
+      "epoch": 0.5451933493476521,
+      "grad_norm": 2.0372812747955322,
+      "learning_rate": 9.100372097404739e-05,
+      "loss": 2.0597,
+      "step": 23150
+    },
+    {
+      "epoch": 0.5454288540341953,
+      "grad_norm": 2.334294319152832,
+      "learning_rate": 9.095662003673874e-05,
+      "loss": 1.9422,
+      "step": 23160
+    },
+    {
+      "epoch": 0.5456643587207386,
+      "grad_norm": 2.039754629135132,
+      "learning_rate": 9.090951909943009e-05,
+      "loss": 2.1631,
+      "step": 23170
+    },
+    {
+      "epoch": 0.5458998634072818,
+      "grad_norm": 1.9584988355636597,
+      "learning_rate": 9.086241816212144e-05,
+      "loss": 2.1069,
+      "step": 23180
+    },
+    {
+      "epoch": 0.546135368093825,
+      "grad_norm": 2.1057446002960205,
+      "learning_rate": 9.081531722481277e-05,
+      "loss": 2.0025,
+      "step": 23190
+    },
+    {
+      "epoch": 0.5463708727803683,
+      "grad_norm": 2.6134400367736816,
+      "learning_rate": 9.076821628750412e-05,
+      "loss": 2.2541,
+      "step": 23200
+    },
+    {
+      "epoch": 0.5466063774669115,
+      "grad_norm": 1.9767898321151733,
+      "learning_rate": 9.072111535019547e-05,
+      "loss": 1.9391,
+      "step": 23210
+    },
+    {
+      "epoch": 0.5468418821534549,
+      "grad_norm": 2.1497464179992676,
+      "learning_rate": 9.067401441288682e-05,
+      "loss": 1.992,
+      "step": 23220
+    },
+    {
+      "epoch": 0.5470773868399981,
+      "grad_norm": 1.8777291774749756,
+      "learning_rate": 9.062691347557816e-05,
+      "loss": 2.0687,
+      "step": 23230
+    },
+    {
+      "epoch": 0.5473128915265414,
+      "grad_norm": 2.7456014156341553,
+      "learning_rate": 9.057981253826951e-05,
+      "loss": 1.9403,
+      "step": 23240
+    },
+    {
+      "epoch": 0.5475483962130846,
+      "grad_norm": 1.9173210859298706,
+      "learning_rate": 9.053271160096086e-05,
+      "loss": 2.362,
+      "step": 23250
+    },
+    {
+      "epoch": 0.5477839008996279,
+      "grad_norm": 3.4497084617614746,
+      "learning_rate": 9.048561066365221e-05,
+      "loss": 2.3116,
+      "step": 23260
+    },
+    {
+      "epoch": 0.5480194055861711,
+      "grad_norm": 1.9495487213134766,
+      "learning_rate": 9.043850972634356e-05,
+      "loss": 2.2498,
+      "step": 23270
+    },
+    {
+      "epoch": 0.5482549102727144,
+      "grad_norm": 2.1822617053985596,
+      "learning_rate": 9.03914087890349e-05,
+      "loss": 2.1444,
+      "step": 23280
+    },
+    {
+      "epoch": 0.5484904149592577,
+      "grad_norm": 2.625450849533081,
+      "learning_rate": 9.034430785172625e-05,
+      "loss": 2.1074,
+      "step": 23290
+    },
+    {
+      "epoch": 0.548725919645801,
+      "grad_norm": 2.899829149246216,
+      "learning_rate": 9.02972069144176e-05,
+      "loss": 2.1126,
+      "step": 23300
+    },
+    {
+      "epoch": 0.5489614243323442,
+      "grad_norm": 3.0588479042053223,
+      "learning_rate": 9.025010597710895e-05,
+      "loss": 1.9202,
+      "step": 23310
+    },
+    {
+      "epoch": 0.5491969290188875,
+      "grad_norm": 2.1747286319732666,
+      "learning_rate": 9.020300503980029e-05,
+      "loss": 1.9893,
+      "step": 23320
+    },
+    {
+      "epoch": 0.5494324337054307,
+      "grad_norm": 2.146778106689453,
+      "learning_rate": 9.015590410249164e-05,
+      "loss": 1.988,
+      "step": 23330
+    },
+    {
+      "epoch": 0.549667938391974,
+      "grad_norm": 1.8845597505569458,
+      "learning_rate": 9.010880316518299e-05,
+      "loss": 2.1452,
+      "step": 23340
+    },
+    {
+      "epoch": 0.5499034430785172,
+      "grad_norm": 2.016979694366455,
+      "learning_rate": 9.006170222787434e-05,
+      "loss": 2.1675,
+      "step": 23350
+    },
+    {
+      "epoch": 0.5501389477650606,
+      "grad_norm": 2.911409378051758,
+      "learning_rate": 9.001460129056569e-05,
+      "loss": 2.0028,
+      "step": 23360
+    },
+    {
+      "epoch": 0.5503744524516038,
+      "grad_norm": 2.8870551586151123,
+      "learning_rate": 8.996750035325704e-05,
+      "loss": 1.9068,
+      "step": 23370
+    },
+    {
+      "epoch": 0.5506099571381471,
+      "grad_norm": 2.1140267848968506,
+      "learning_rate": 8.992039941594839e-05,
+      "loss": 2.1994,
+      "step": 23380
+    },
+    {
+      "epoch": 0.5508454618246903,
+      "grad_norm": 1.9857615232467651,
+      "learning_rate": 8.987329847863974e-05,
+      "loss": 2.1649,
+      "step": 23390
+    },
+    {
+      "epoch": 0.5510809665112336,
+      "grad_norm": 2.7637999057769775,
+      "learning_rate": 8.982619754133107e-05,
+      "loss": 1.9992,
+      "step": 23400
+    },
+    {
+      "epoch": 0.5513164711977768,
+      "grad_norm": 2.225193738937378,
+      "learning_rate": 8.977909660402243e-05,
+      "loss": 2.1342,
+      "step": 23410
+    },
+    {
+      "epoch": 0.5515519758843201,
+      "grad_norm": 1.7296162843704224,
+      "learning_rate": 8.973199566671378e-05,
+      "loss": 2.0405,
+      "step": 23420
+    },
+    {
+      "epoch": 0.5517874805708634,
+      "grad_norm": 2.2399344444274902,
+      "learning_rate": 8.968489472940513e-05,
+      "loss": 2.0995,
+      "step": 23430
+    },
+    {
+      "epoch": 0.5520229852574067,
+      "grad_norm": 2.2582974433898926,
+      "learning_rate": 8.963779379209648e-05,
+      "loss": 1.9908,
+      "step": 23440
+    },
+    {
+      "epoch": 0.5522584899439499,
+      "grad_norm": 2.6607015132904053,
+      "learning_rate": 8.959069285478781e-05,
+      "loss": 1.9292,
+      "step": 23450
+    },
+    {
+      "epoch": 0.5524939946304932,
+      "grad_norm": 2.3309760093688965,
+      "learning_rate": 8.954359191747916e-05,
+      "loss": 1.874,
+      "step": 23460
+    },
+    {
+      "epoch": 0.5527294993170364,
+      "grad_norm": 2.2204902172088623,
+      "learning_rate": 8.949649098017051e-05,
+      "loss": 2.064,
+      "step": 23470
+    },
+    {
+      "epoch": 0.5529650040035796,
+      "grad_norm": 1.4835526943206787,
+      "learning_rate": 8.944939004286186e-05,
+      "loss": 1.9785,
+      "step": 23480
+    },
+    {
+      "epoch": 0.5532005086901229,
+      "grad_norm": 2.3408985137939453,
+      "learning_rate": 8.94022891055532e-05,
+      "loss": 2.0047,
+      "step": 23490
+    },
+    {
+      "epoch": 0.5534360133766661,
+      "grad_norm": 2.300576686859131,
+      "learning_rate": 8.935518816824455e-05,
+      "loss": 2.1159,
+      "step": 23500
+    },
+    {
+      "epoch": 0.5536715180632095,
+      "grad_norm": 1.9129767417907715,
+      "learning_rate": 8.93080872309359e-05,
+      "loss": 2.0221,
+      "step": 23510
+    },
+    {
+      "epoch": 0.5539070227497527,
+      "grad_norm": 1.8746204376220703,
+      "learning_rate": 8.926098629362725e-05,
+      "loss": 2.1288,
+      "step": 23520
+    },
+    {
+      "epoch": 0.554142527436296,
+      "grad_norm": 1.7646563053131104,
+      "learning_rate": 8.921388535631859e-05,
+      "loss": 2.2262,
+      "step": 23530
+    },
+    {
+      "epoch": 0.5543780321228392,
+      "grad_norm": 1.9448425769805908,
+      "learning_rate": 8.916678441900994e-05,
+      "loss": 1.8702,
+      "step": 23540
+    },
+    {
+      "epoch": 0.5546135368093825,
+      "grad_norm": 2.126228094100952,
+      "learning_rate": 8.911968348170129e-05,
+      "loss": 2.1124,
+      "step": 23550
+    },
+    {
+      "epoch": 0.5548490414959257,
+      "grad_norm": 1.686448097229004,
+      "learning_rate": 8.907258254439264e-05,
+      "loss": 2.0523,
+      "step": 23560
+    },
+    {
+      "epoch": 0.555084546182469,
+      "grad_norm": 2.2924816608428955,
+      "learning_rate": 8.902548160708399e-05,
+      "loss": 2.1922,
+      "step": 23570
+    },
+    {
+      "epoch": 0.5553200508690123,
+      "grad_norm": 2.001551866531372,
+      "learning_rate": 8.897838066977532e-05,
+      "loss": 1.8586,
+      "step": 23580
+    },
+    {
+      "epoch": 0.5555555555555556,
+      "grad_norm": 2.2796730995178223,
+      "learning_rate": 8.893127973246668e-05,
+      "loss": 2.1119,
+      "step": 23590
+    },
+    {
+      "epoch": 0.5557910602420988,
+      "grad_norm": 1.9059703350067139,
+      "learning_rate": 8.888417879515803e-05,
+      "loss": 2.1778,
+      "step": 23600
+    },
+    {
+      "epoch": 0.5560265649286421,
+      "grad_norm": 2.31387996673584,
+      "learning_rate": 8.883707785784938e-05,
+      "loss": 1.9897,
+      "step": 23610
+    },
+    {
+      "epoch": 0.5562620696151853,
+      "grad_norm": 2.014240264892578,
+      "learning_rate": 8.878997692054071e-05,
+      "loss": 2.12,
+      "step": 23620
+    },
+    {
+      "epoch": 0.5564975743017286,
+      "grad_norm": 2.1083483695983887,
+      "learning_rate": 8.874287598323206e-05,
+      "loss": 1.9658,
+      "step": 23630
+    },
+    {
+      "epoch": 0.5567330789882718,
+      "grad_norm": 1.9519131183624268,
+      "learning_rate": 8.869577504592341e-05,
+      "loss": 2.2165,
+      "step": 23640
+    },
+    {
+      "epoch": 0.5569685836748152,
+      "grad_norm": 1.703078031539917,
+      "learning_rate": 8.864867410861478e-05,
+      "loss": 2.1628,
+      "step": 23650
+    },
+    {
+      "epoch": 0.5572040883613584,
+      "grad_norm": 2.055586814880371,
+      "learning_rate": 8.860157317130611e-05,
+      "loss": 1.9498,
+      "step": 23660
+    },
+    {
+      "epoch": 0.5574395930479017,
+      "grad_norm": 2.0325064659118652,
+      "learning_rate": 8.855447223399746e-05,
+      "loss": 1.9808,
+      "step": 23670
+    },
+    {
+      "epoch": 0.5576750977344449,
+      "grad_norm": 1.8139878511428833,
+      "learning_rate": 8.850737129668881e-05,
+      "loss": 2.0656,
+      "step": 23680
+    },
+    {
+      "epoch": 0.5579106024209882,
+      "grad_norm": 2.5634472370147705,
+      "learning_rate": 8.846027035938016e-05,
+      "loss": 1.9946,
+      "step": 23690
+    },
+    {
+      "epoch": 0.5581461071075314,
+      "grad_norm": 2.3983256816864014,
+      "learning_rate": 8.84131694220715e-05,
+      "loss": 2.0557,
+      "step": 23700
+    },
+    {
+      "epoch": 0.5583816117940748,
+      "grad_norm": 1.7592830657958984,
+      "learning_rate": 8.836606848476285e-05,
+      "loss": 1.9625,
+      "step": 23710
+    },
+    {
+      "epoch": 0.558617116480618,
+      "grad_norm": 1.9398764371871948,
+      "learning_rate": 8.83189675474542e-05,
+      "loss": 1.9973,
+      "step": 23720
+    },
+    {
+      "epoch": 0.5588526211671613,
+      "grad_norm": 2.4670112133026123,
+      "learning_rate": 8.827186661014555e-05,
+      "loss": 2.0338,
+      "step": 23730
+    },
+    {
+      "epoch": 0.5590881258537045,
+      "grad_norm": 2.532526969909668,
+      "learning_rate": 8.82247656728369e-05,
+      "loss": 2.0994,
+      "step": 23740
+    },
+    {
+      "epoch": 0.5593236305402477,
+      "grad_norm": 3.6450741291046143,
+      "learning_rate": 8.81823748292591e-05,
+      "loss": 2.1508,
+      "step": 23750
+    },
+    {
+      "epoch": 0.559559135226791,
+      "grad_norm": 2.395998239517212,
+      "learning_rate": 8.813527389195046e-05,
+      "loss": 2.035,
+      "step": 23760
+    },
+    {
+      "epoch": 0.5597946399133342,
+      "grad_norm": 2.092289686203003,
+      "learning_rate": 8.80881729546418e-05,
+      "loss": 2.0636,
+      "step": 23770
+    },
+    {
+      "epoch": 0.5600301445998775,
+      "grad_norm": 1.8740358352661133,
+      "learning_rate": 8.804107201733314e-05,
+      "loss": 1.9921,
+      "step": 23780
+    },
+    {
+      "epoch": 0.5602656492864208,
+      "grad_norm": 3.254704475402832,
+      "learning_rate": 8.79939710800245e-05,
+      "loss": 2.0609,
+      "step": 23790
+    },
+    {
+      "epoch": 0.5605011539729641,
+      "grad_norm": 2.3679122924804688,
+      "learning_rate": 8.794687014271584e-05,
+      "loss": 1.9789,
+      "step": 23800
+    },
+    {
+      "epoch": 0.5607366586595073,
+      "grad_norm": 2.638920783996582,
+      "learning_rate": 8.78997692054072e-05,
+      "loss": 2.0542,
+      "step": 23810
+    },
+    {
+      "epoch": 0.5609721633460506,
+      "grad_norm": 2.1278998851776123,
+      "learning_rate": 8.785266826809853e-05,
+      "loss": 2.185,
+      "step": 23820
+    },
+    {
+      "epoch": 0.5612076680325938,
+      "grad_norm": 2.04278826713562,
+      "learning_rate": 8.780556733078988e-05,
+      "loss": 2.1442,
+      "step": 23830
+    },
+    {
+      "epoch": 0.5614431727191371,
+      "grad_norm": 2.327648639678955,
+      "learning_rate": 8.775846639348123e-05,
+      "loss": 2.2638,
+      "step": 23840
+    },
+    {
+      "epoch": 0.5616786774056803,
+      "grad_norm": 2.190995454788208,
+      "learning_rate": 8.771136545617258e-05,
+      "loss": 2.3052,
+      "step": 23850
+    },
+    {
+      "epoch": 0.5619141820922237,
+      "grad_norm": 2.5929126739501953,
+      "learning_rate": 8.766426451886393e-05,
+      "loss": 2.078,
+      "step": 23860
+    },
+    {
+      "epoch": 0.5621496867787669,
+      "grad_norm": 2.1054494380950928,
+      "learning_rate": 8.761716358155527e-05,
+      "loss": 2.1903,
+      "step": 23870
+    },
+    {
+      "epoch": 0.5623851914653102,
+      "grad_norm": 2.804368495941162,
+      "learning_rate": 8.757006264424662e-05,
+      "loss": 1.8636,
+      "step": 23880
+    },
+    {
+      "epoch": 0.5626206961518534,
+      "grad_norm": 2.6225225925445557,
+      "learning_rate": 8.752296170693797e-05,
+      "loss": 2.0436,
+      "step": 23890
+    },
+    {
+      "epoch": 0.5628562008383967,
+      "grad_norm": 1.7755427360534668,
+      "learning_rate": 8.747586076962932e-05,
+      "loss": 2.1628,
+      "step": 23900
+    },
+    {
+      "epoch": 0.5630917055249399,
+      "grad_norm": 2.0387985706329346,
+      "learning_rate": 8.742875983232067e-05,
+      "loss": 2.1442,
+      "step": 23910
+    },
+    {
+      "epoch": 0.5633272102114832,
+      "grad_norm": 2.0750441551208496,
+      "learning_rate": 8.738165889501202e-05,
+      "loss": 2.0374,
+      "step": 23920
+    },
+    {
+      "epoch": 0.5635627148980265,
+      "grad_norm": 2.1440858840942383,
+      "learning_rate": 8.733455795770337e-05,
+      "loss": 2.0789,
+      "step": 23930
+    },
+    {
+      "epoch": 0.5637982195845698,
+      "grad_norm": 2.1340370178222656,
+      "learning_rate": 8.728745702039472e-05,
+      "loss": 2.2198,
+      "step": 23940
+    },
+    {
+      "epoch": 0.564033724271113,
+      "grad_norm": 2.3632407188415527,
+      "learning_rate": 8.724035608308606e-05,
+      "loss": 2.0647,
+      "step": 23950
+    },
+    {
+      "epoch": 0.5642692289576563,
+      "grad_norm": 2.03859543800354,
+      "learning_rate": 8.719325514577741e-05,
+      "loss": 2.1071,
+      "step": 23960
+    },
+    {
+      "epoch": 0.5645047336441995,
+      "grad_norm": 2.2486722469329834,
+      "learning_rate": 8.714615420846876e-05,
+      "loss": 2.0443,
+      "step": 23970
+    },
+    {
+      "epoch": 0.5647402383307428,
+      "grad_norm": 1.8128294944763184,
+      "learning_rate": 8.709905327116011e-05,
+      "loss": 1.9782,
+      "step": 23980
+    },
+    {
+      "epoch": 0.564975743017286,
+      "grad_norm": 2.3327388763427734,
+      "learning_rate": 8.705195233385144e-05,
+      "loss": 2.1579,
+      "step": 23990
+    },
+    {
+      "epoch": 0.5652112477038294,
+      "grad_norm": 2.0959858894348145,
+      "learning_rate": 8.70048513965428e-05,
+      "loss": 1.8995,
+      "step": 24000
+    },
+    {
+      "epoch": 0.5654467523903726,
+      "grad_norm": 2.87188458442688,
+      "learning_rate": 8.695775045923414e-05,
+      "loss": 2.0005,
+      "step": 24010
+    },
+    {
+      "epoch": 0.5656822570769159,
+      "grad_norm": 2.495626449584961,
+      "learning_rate": 8.69106495219255e-05,
+      "loss": 2.3682,
+      "step": 24020
+    },
+    {
+      "epoch": 0.5659177617634591,
+      "grad_norm": 3.239908456802368,
+      "learning_rate": 8.686354858461683e-05,
+      "loss": 1.9192,
+      "step": 24030
+    },
+    {
+      "epoch": 0.5661532664500023,
+      "grad_norm": 1.99828040599823,
+      "learning_rate": 8.681644764730818e-05,
+      "loss": 2.3154,
+      "step": 24040
+    },
+    {
+      "epoch": 0.5663887711365456,
+      "grad_norm": 2.1217734813690186,
+      "learning_rate": 8.676934670999953e-05,
+      "loss": 1.9546,
+      "step": 24050
+    },
+    {
+      "epoch": 0.5666242758230888,
+      "grad_norm": 3.101858377456665,
+      "learning_rate": 8.672224577269088e-05,
+      "loss": 2.0897,
+      "step": 24060
+    },
+    {
+      "epoch": 0.5668597805096321,
+      "grad_norm": 2.358698606491089,
+      "learning_rate": 8.667514483538223e-05,
+      "loss": 2.0906,
+      "step": 24070
+    },
+    {
+      "epoch": 0.5670952851961754,
+      "grad_norm": 2.2802388668060303,
+      "learning_rate": 8.662804389807357e-05,
+      "loss": 2.0595,
+      "step": 24080
+    },
+    {
+      "epoch": 0.5673307898827187,
+      "grad_norm": 2.175506114959717,
+      "learning_rate": 8.658094296076492e-05,
+      "loss": 1.8499,
+      "step": 24090
+    },
+    {
+      "epoch": 0.5675662945692619,
+      "grad_norm": 1.782609224319458,
+      "learning_rate": 8.653384202345627e-05,
+      "loss": 2.1053,
+      "step": 24100
+    },
+    {
+      "epoch": 0.5678017992558052,
+      "grad_norm": 1.9120488166809082,
+      "learning_rate": 8.648674108614762e-05,
+      "loss": 2.2141,
+      "step": 24110
+    },
+    {
+      "epoch": 0.5680373039423484,
+      "grad_norm": 3.216996908187866,
+      "learning_rate": 8.643964014883896e-05,
+      "loss": 2.0973,
+      "step": 24120
+    },
+    {
+      "epoch": 0.5682728086288917,
+      "grad_norm": 2.160984754562378,
+      "learning_rate": 8.639253921153031e-05,
+      "loss": 2.1362,
+      "step": 24130
+    },
+    {
+      "epoch": 0.5685083133154349,
+      "grad_norm": 1.6789969205856323,
+      "learning_rate": 8.634543827422166e-05,
+      "loss": 2.1757,
+      "step": 24140
+    },
+    {
+      "epoch": 0.5687438180019783,
+      "grad_norm": 2.3847877979278564,
+      "learning_rate": 8.629833733691301e-05,
+      "loss": 2.2551,
+      "step": 24150
+    },
+    {
+      "epoch": 0.5689793226885215,
+      "grad_norm": 2.110700845718384,
+      "learning_rate": 8.625123639960436e-05,
+      "loss": 2.1488,
+      "step": 24160
+    },
+    {
+      "epoch": 0.5692148273750648,
+      "grad_norm": 2.052661657333374,
+      "learning_rate": 8.62041354622957e-05,
+      "loss": 2.166,
+      "step": 24170
+    },
+    {
+      "epoch": 0.569450332061608,
+      "grad_norm": 2.335162401199341,
+      "learning_rate": 8.615703452498704e-05,
+      "loss": 1.9717,
+      "step": 24180
+    },
+    {
+      "epoch": 0.5696858367481513,
+      "grad_norm": 3.1710004806518555,
+      "learning_rate": 8.610993358767841e-05,
+      "loss": 1.8111,
+      "step": 24190
+    },
+    {
+      "epoch": 0.5699213414346945,
+      "grad_norm": 2.3458499908447266,
+      "learning_rate": 8.606283265036975e-05,
+      "loss": 2.159,
+      "step": 24200
+    },
+    {
+      "epoch": 0.5701568461212378,
+      "grad_norm": 2.641172170639038,
+      "learning_rate": 8.60157317130611e-05,
+      "loss": 1.9753,
+      "step": 24210
+    },
+    {
+      "epoch": 0.5703923508077811,
+      "grad_norm": 1.8560911417007446,
+      "learning_rate": 8.596863077575245e-05,
+      "loss": 2.1409,
+      "step": 24220
+    },
+    {
+      "epoch": 0.5706278554943244,
+      "grad_norm": 2.3327791690826416,
+      "learning_rate": 8.59215298384438e-05,
+      "loss": 2.0098,
+      "step": 24230
+    },
+    {
+      "epoch": 0.5708633601808676,
+      "grad_norm": 2.3636698722839355,
+      "learning_rate": 8.587442890113515e-05,
+      "loss": 2.1375,
+      "step": 24240
+    },
+    {
+      "epoch": 0.5710988648674109,
+      "grad_norm": 2.747985601425171,
+      "learning_rate": 8.582732796382648e-05,
+      "loss": 2.1451,
+      "step": 24250
+    },
+    {
+      "epoch": 0.5713343695539541,
+      "grad_norm": 2.1141562461853027,
+      "learning_rate": 8.578022702651783e-05,
+      "loss": 2.1145,
+      "step": 24260
+    },
+    {
+      "epoch": 0.5715698742404974,
+      "grad_norm": 2.0584163665771484,
+      "learning_rate": 8.573312608920918e-05,
+      "loss": 1.919,
+      "step": 24270
+    },
+    {
+      "epoch": 0.5718053789270406,
+      "grad_norm": 2.1240551471710205,
+      "learning_rate": 8.568602515190053e-05,
+      "loss": 2.0322,
+      "step": 24280
+    },
+    {
+      "epoch": 0.572040883613584,
+      "grad_norm": 2.4911723136901855,
+      "learning_rate": 8.563892421459187e-05,
+      "loss": 2.4528,
+      "step": 24290
+    },
+    {
+      "epoch": 0.5722763883001272,
+      "grad_norm": 1.859668493270874,
+      "learning_rate": 8.559182327728322e-05,
+      "loss": 2.027,
+      "step": 24300
+    },
+    {
+      "epoch": 0.5725118929866705,
+      "grad_norm": 1.7338566780090332,
+      "learning_rate": 8.554472233997457e-05,
+      "loss": 1.9711,
+      "step": 24310
+    },
+    {
+      "epoch": 0.5727473976732137,
+      "grad_norm": 2.1398377418518066,
+      "learning_rate": 8.549762140266592e-05,
+      "loss": 2.1034,
+      "step": 24320
+    },
+    {
+      "epoch": 0.5729829023597569,
+      "grad_norm": 1.8699679374694824,
+      "learning_rate": 8.545052046535726e-05,
+      "loss": 1.9032,
+      "step": 24330
+    },
+    {
+      "epoch": 0.5732184070463002,
+      "grad_norm": 2.045719623565674,
+      "learning_rate": 8.540341952804861e-05,
+      "loss": 1.9111,
+      "step": 24340
+    },
+    {
+      "epoch": 0.5734539117328434,
+      "grad_norm": 2.614588499069214,
+      "learning_rate": 8.535631859073996e-05,
+      "loss": 2.0196,
+      "step": 24350
+    },
+    {
+      "epoch": 0.5736894164193868,
+      "grad_norm": 2.135077953338623,
+      "learning_rate": 8.530921765343131e-05,
+      "loss": 2.2211,
+      "step": 24360
+    },
+    {
+      "epoch": 0.57392492110593,
+      "grad_norm": 2.3194169998168945,
+      "learning_rate": 8.526211671612266e-05,
+      "loss": 2.1909,
+      "step": 24370
+    },
+    {
+      "epoch": 0.5741604257924733,
+      "grad_norm": 2.06208872795105,
+      "learning_rate": 8.5215015778814e-05,
+      "loss": 2.099,
+      "step": 24380
+    },
+    {
+      "epoch": 0.5743959304790165,
+      "grad_norm": 1.9081873893737793,
+      "learning_rate": 8.516791484150535e-05,
+      "loss": 1.8989,
+      "step": 24390
+    },
+    {
+      "epoch": 0.5746314351655598,
+      "grad_norm": 1.925196886062622,
+      "learning_rate": 8.51208139041967e-05,
+      "loss": 2.1537,
+      "step": 24400
+    },
+    {
+      "epoch": 0.574866939852103,
+      "grad_norm": 2.0393810272216797,
+      "learning_rate": 8.507371296688805e-05,
+      "loss": 1.9381,
+      "step": 24410
+    },
+    {
+      "epoch": 0.5751024445386463,
+      "grad_norm": 2.061901807785034,
+      "learning_rate": 8.502661202957938e-05,
+      "loss": 2.0184,
+      "step": 24420
+    },
+    {
+      "epoch": 0.5753379492251895,
+      "grad_norm": 1.582486629486084,
+      "learning_rate": 8.497951109227073e-05,
+      "loss": 2.1767,
+      "step": 24430
+    },
+    {
+      "epoch": 0.5755734539117329,
+      "grad_norm": 2.24311900138855,
+      "learning_rate": 8.493241015496208e-05,
+      "loss": 2.1155,
+      "step": 24440
+    },
+    {
+      "epoch": 0.5758089585982761,
+      "grad_norm": 1.794895052909851,
+      "learning_rate": 8.488530921765343e-05,
+      "loss": 1.8587,
+      "step": 24450
+    },
+    {
+      "epoch": 0.5760444632848194,
+      "grad_norm": 2.3999338150024414,
+      "learning_rate": 8.483820828034478e-05,
+      "loss": 2.1511,
+      "step": 24460
+    },
+    {
+      "epoch": 0.5762799679713626,
+      "grad_norm": 2.538761854171753,
+      "learning_rate": 8.479110734303613e-05,
+      "loss": 2.1788,
+      "step": 24470
+    },
+    {
+      "epoch": 0.5765154726579059,
+      "grad_norm": 2.7868943214416504,
+      "learning_rate": 8.474400640572748e-05,
+      "loss": 2.0722,
+      "step": 24480
+    },
+    {
+      "epoch": 0.5767509773444491,
+      "grad_norm": 2.052523136138916,
+      "learning_rate": 8.469690546841883e-05,
+      "loss": 2.1654,
+      "step": 24490
+    },
+    {
+      "epoch": 0.5769864820309925,
+      "grad_norm": 2.1687915325164795,
+      "learning_rate": 8.464980453111017e-05,
+      "loss": 1.9479,
+      "step": 24500
+    },
+    {
+      "epoch": 0.5772219867175357,
+      "grad_norm": 2.0487418174743652,
+      "learning_rate": 8.460270359380152e-05,
+      "loss": 2.1456,
+      "step": 24510
+    },
+    {
+      "epoch": 0.577457491404079,
+      "grad_norm": 2.201139211654663,
+      "learning_rate": 8.455560265649287e-05,
+      "loss": 2.0357,
+      "step": 24520
+    },
+    {
+      "epoch": 0.5776929960906222,
+      "grad_norm": 2.2305378913879395,
+      "learning_rate": 8.450850171918422e-05,
+      "loss": 2.1552,
+      "step": 24530
+    },
+    {
+      "epoch": 0.5779285007771655,
+      "grad_norm": 4.121927261352539,
+      "learning_rate": 8.446140078187557e-05,
+      "loss": 2.0478,
+      "step": 24540
+    },
+    {
+      "epoch": 0.5781640054637087,
+      "grad_norm": 2.2727227210998535,
+      "learning_rate": 8.441429984456691e-05,
+      "loss": 1.9688,
+      "step": 24550
+    },
+    {
+      "epoch": 0.578399510150252,
+      "grad_norm": 2.3096418380737305,
+      "learning_rate": 8.436719890725826e-05,
+      "loss": 2.0188,
+      "step": 24560
+    },
+    {
+      "epoch": 0.5786350148367952,
+      "grad_norm": 1.787701964378357,
+      "learning_rate": 8.432009796994961e-05,
+      "loss": 1.9433,
+      "step": 24570
+    },
+    {
+      "epoch": 0.5788705195233386,
+      "grad_norm": 1.5643645524978638,
+      "learning_rate": 8.427299703264096e-05,
+      "loss": 1.9101,
+      "step": 24580
+    },
+    {
+      "epoch": 0.5791060242098818,
+      "grad_norm": 2.1164591312408447,
+      "learning_rate": 8.42258960953323e-05,
+      "loss": 2.0142,
+      "step": 24590
+    },
+    {
+      "epoch": 0.579341528896425,
+      "grad_norm": 1.9861196279525757,
+      "learning_rate": 8.417879515802365e-05,
+      "loss": 2.1013,
+      "step": 24600
+    },
+    {
+      "epoch": 0.5795770335829683,
+      "grad_norm": 2.388425588607788,
+      "learning_rate": 8.4131694220715e-05,
+      "loss": 1.9681,
+      "step": 24610
+    },
+    {
+      "epoch": 0.5798125382695115,
+      "grad_norm": 1.6500285863876343,
+      "learning_rate": 8.408459328340635e-05,
+      "loss": 1.992,
+      "step": 24620
+    },
+    {
+      "epoch": 0.5800480429560548,
+      "grad_norm": 1.7825500965118408,
+      "learning_rate": 8.403749234609768e-05,
+      "loss": 2.0607,
+      "step": 24630
+    },
+    {
+      "epoch": 0.580283547642598,
+      "grad_norm": 2.313638210296631,
+      "learning_rate": 8.399039140878903e-05,
+      "loss": 2.0777,
+      "step": 24640
+    },
+    {
+      "epoch": 0.5805190523291414,
+      "grad_norm": 2.432238817214966,
+      "learning_rate": 8.394329047148038e-05,
+      "loss": 2.0948,
+      "step": 24650
+    },
+    {
+      "epoch": 0.5807545570156846,
+      "grad_norm": 2.8354523181915283,
+      "learning_rate": 8.389618953417173e-05,
+      "loss": 2.1771,
+      "step": 24660
+    },
+    {
+      "epoch": 0.5809900617022279,
+      "grad_norm": 2.3798136711120605,
+      "learning_rate": 8.384908859686308e-05,
+      "loss": 1.9815,
+      "step": 24670
+    },
+    {
+      "epoch": 0.5812255663887711,
+      "grad_norm": 2.092031478881836,
+      "learning_rate": 8.380198765955442e-05,
+      "loss": 2.0798,
+      "step": 24680
+    },
+    {
+      "epoch": 0.5814610710753144,
+      "grad_norm": 2.465062379837036,
+      "learning_rate": 8.375488672224577e-05,
+      "loss": 1.8865,
+      "step": 24690
+    },
+    {
+      "epoch": 0.5816965757618576,
+      "grad_norm": 2.506052017211914,
+      "learning_rate": 8.370778578493712e-05,
+      "loss": 1.9352,
+      "step": 24700
+    },
+    {
+      "epoch": 0.581932080448401,
+      "grad_norm": 2.2928619384765625,
+      "learning_rate": 8.366068484762847e-05,
+      "loss": 1.8513,
+      "step": 24710
+    },
+    {
+      "epoch": 0.5821675851349442,
+      "grad_norm": 1.858624815940857,
+      "learning_rate": 8.361358391031981e-05,
+      "loss": 2.095,
+      "step": 24720
+    },
+    {
+      "epoch": 0.5824030898214875,
+      "grad_norm": 2.981670379638672,
+      "learning_rate": 8.356648297301116e-05,
+      "loss": 1.978,
+      "step": 24730
+    },
+    {
+      "epoch": 0.5826385945080307,
+      "grad_norm": 2.1256954669952393,
+      "learning_rate": 8.351938203570251e-05,
+      "loss": 1.9985,
+      "step": 24740
+    },
+    {
+      "epoch": 0.582874099194574,
+      "grad_norm": 1.9734386205673218,
+      "learning_rate": 8.347228109839387e-05,
+      "loss": 2.0097,
+      "step": 24750
+    },
+    {
+      "epoch": 0.5831096038811172,
+      "grad_norm": 2.111104965209961,
+      "learning_rate": 8.342518016108521e-05,
+      "loss": 2.0503,
+      "step": 24760
+    },
+    {
+      "epoch": 0.5833451085676605,
+      "grad_norm": 1.7222291231155396,
+      "learning_rate": 8.337807922377656e-05,
+      "loss": 1.8505,
+      "step": 24770
+    },
+    {
+      "epoch": 0.5835806132542037,
+      "grad_norm": 2.5214107036590576,
+      "learning_rate": 8.333097828646791e-05,
+      "loss": 2.0998,
+      "step": 24780
+    },
+    {
+      "epoch": 0.5838161179407471,
+      "grad_norm": 2.7706222534179688,
+      "learning_rate": 8.328387734915926e-05,
+      "loss": 2.0881,
+      "step": 24790
+    },
+    {
+      "epoch": 0.5840516226272903,
+      "grad_norm": 2.2346792221069336,
+      "learning_rate": 8.32367764118506e-05,
+      "loss": 2.0906,
+      "step": 24800
+    },
+    {
+      "epoch": 0.5842871273138336,
+      "grad_norm": 2.1181459426879883,
+      "learning_rate": 8.318967547454195e-05,
+      "loss": 1.9335,
+      "step": 24810
+    },
+    {
+      "epoch": 0.5845226320003768,
+      "grad_norm": 2.1477532386779785,
+      "learning_rate": 8.31425745372333e-05,
+      "loss": 2.1249,
+      "step": 24820
+    },
+    {
+      "epoch": 0.5847581366869201,
+      "grad_norm": 2.5332417488098145,
+      "learning_rate": 8.309547359992465e-05,
+      "loss": 2.1898,
+      "step": 24830
+    },
+    {
+      "epoch": 0.5849936413734633,
+      "grad_norm": 1.6639565229415894,
+      "learning_rate": 8.3048372662616e-05,
+      "loss": 2.077,
+      "step": 24840
+    },
+    {
+      "epoch": 0.5852291460600066,
+      "grad_norm": 2.576613187789917,
+      "learning_rate": 8.300127172530733e-05,
+      "loss": 2.1235,
+      "step": 24850
+    },
+    {
+      "epoch": 0.5854646507465499,
+      "grad_norm": 2.24748158454895,
+      "learning_rate": 8.295417078799869e-05,
+      "loss": 1.7241,
+      "step": 24860
+    },
+    {
+      "epoch": 0.5857001554330932,
+      "grad_norm": 2.139662265777588,
+      "learning_rate": 8.290706985069004e-05,
+      "loss": 2.0476,
+      "step": 24870
+    },
+    {
+      "epoch": 0.5859356601196364,
+      "grad_norm": 2.513113260269165,
+      "learning_rate": 8.285996891338139e-05,
+      "loss": 2.0448,
+      "step": 24880
+    },
+    {
+      "epoch": 0.5861711648061796,
+      "grad_norm": 2.219991445541382,
+      "learning_rate": 8.281286797607272e-05,
+      "loss": 1.7964,
+      "step": 24890
+    },
+    {
+      "epoch": 0.5864066694927229,
+      "grad_norm": 2.668653964996338,
+      "learning_rate": 8.276576703876407e-05,
+      "loss": 2.2925,
+      "step": 24900
+    },
+    {
+      "epoch": 0.5866421741792661,
+      "grad_norm": 2.3607492446899414,
+      "learning_rate": 8.271866610145542e-05,
+      "loss": 2.0662,
+      "step": 24910
+    },
+    {
+      "epoch": 0.5868776788658094,
+      "grad_norm": 2.165482521057129,
+      "learning_rate": 8.267156516414677e-05,
+      "loss": 2.0108,
+      "step": 24920
+    },
+    {
+      "epoch": 0.5871131835523526,
+      "grad_norm": 3.355048418045044,
+      "learning_rate": 8.262446422683811e-05,
+      "loss": 2.2239,
+      "step": 24930
+    },
+    {
+      "epoch": 0.587348688238896,
+      "grad_norm": 2.3312196731567383,
+      "learning_rate": 8.257736328952946e-05,
+      "loss": 2.1449,
+      "step": 24940
+    },
+    {
+      "epoch": 0.5875841929254392,
+      "grad_norm": 2.766026735305786,
+      "learning_rate": 8.253026235222081e-05,
+      "loss": 2.0531,
+      "step": 24950
+    },
+    {
+      "epoch": 0.5878196976119825,
+      "grad_norm": 2.3361899852752686,
+      "learning_rate": 8.248316141491216e-05,
+      "loss": 2.029,
+      "step": 24960
+    },
+    {
+      "epoch": 0.5880552022985257,
+      "grad_norm": 2.563586950302124,
+      "learning_rate": 8.243606047760351e-05,
+      "loss": 2.0073,
+      "step": 24970
+    },
+    {
+      "epoch": 0.588290706985069,
+      "grad_norm": 3.044583320617676,
+      "learning_rate": 8.238895954029485e-05,
+      "loss": 2.3071,
+      "step": 24980
+    },
+    {
+      "epoch": 0.5885262116716122,
+      "grad_norm": 1.7707631587982178,
+      "learning_rate": 8.23418586029862e-05,
+      "loss": 1.9584,
+      "step": 24990
+    },
+    {
+      "epoch": 0.5887617163581556,
+      "grad_norm": 3.5477962493896484,
+      "learning_rate": 8.229475766567755e-05,
+      "loss": 2.0358,
+      "step": 25000
+    },
+    {
+      "epoch": 0.5889972210446988,
+      "grad_norm": 2.394061326980591,
+      "learning_rate": 8.22476567283689e-05,
+      "loss": 2.0315,
+      "step": 25010
+    },
+    {
+      "epoch": 0.5892327257312421,
+      "grad_norm": 2.087583065032959,
+      "learning_rate": 8.220055579106023e-05,
+      "loss": 2.0567,
+      "step": 25020
+    },
+    {
+      "epoch": 0.5894682304177853,
+      "grad_norm": 1.9279582500457764,
+      "learning_rate": 8.21534548537516e-05,
+      "loss": 2.2003,
+      "step": 25030
+    },
+    {
+      "epoch": 0.5897037351043286,
+      "grad_norm": 2.105403184890747,
+      "learning_rate": 8.210635391644295e-05,
+      "loss": 1.9165,
+      "step": 25040
+    },
+    {
+      "epoch": 0.5899392397908718,
+      "grad_norm": 2.1151294708251953,
+      "learning_rate": 8.20592529791343e-05,
+      "loss": 2.1332,
+      "step": 25050
+    },
+    {
+      "epoch": 0.5901747444774151,
+      "grad_norm": 2.34464693069458,
+      "learning_rate": 8.201215204182564e-05,
+      "loss": 1.9142,
+      "step": 25060
+    },
+    {
+      "epoch": 0.5904102491639583,
+      "grad_norm": 2.0007407665252686,
+      "learning_rate": 8.196505110451699e-05,
+      "loss": 1.9881,
+      "step": 25070
+    },
+    {
+      "epoch": 0.5906457538505017,
+      "grad_norm": 2.4901375770568848,
+      "learning_rate": 8.191795016720834e-05,
+      "loss": 2.0313,
+      "step": 25080
+    },
+    {
+      "epoch": 0.5908812585370449,
+      "grad_norm": 1.5879149436950684,
+      "learning_rate": 8.187084922989969e-05,
+      "loss": 2.0404,
+      "step": 25090
+    },
+    {
+      "epoch": 0.5911167632235882,
+      "grad_norm": 1.9847928285598755,
+      "learning_rate": 8.182374829259102e-05,
+      "loss": 1.9612,
+      "step": 25100
+    },
+    {
+      "epoch": 0.5913522679101314,
+      "grad_norm": 1.7049063444137573,
+      "learning_rate": 8.177664735528237e-05,
+      "loss": 2.1503,
+      "step": 25110
+    },
+    {
+      "epoch": 0.5915877725966747,
+      "grad_norm": 2.089395046234131,
+      "learning_rate": 8.172954641797372e-05,
+      "loss": 2.1733,
+      "step": 25120
+    },
+    {
+      "epoch": 0.5918232772832179,
+      "grad_norm": 2.2690000534057617,
+      "learning_rate": 8.168244548066507e-05,
+      "loss": 2.1087,
+      "step": 25130
+    },
+    {
+      "epoch": 0.5920587819697612,
+      "grad_norm": 2.3506076335906982,
+      "learning_rate": 8.163534454335642e-05,
+      "loss": 1.9857,
+      "step": 25140
+    },
+    {
+      "epoch": 0.5922942866563045,
+      "grad_norm": 2.1546356678009033,
+      "learning_rate": 8.158824360604776e-05,
+      "loss": 1.8125,
+      "step": 25150
+    },
+    {
+      "epoch": 0.5925297913428478,
+      "grad_norm": 2.6273109912872314,
+      "learning_rate": 8.154114266873911e-05,
+      "loss": 2.1505,
+      "step": 25160
+    },
+    {
+      "epoch": 0.592765296029391,
+      "grad_norm": 2.0121755599975586,
+      "learning_rate": 8.149404173143046e-05,
+      "loss": 2.2848,
+      "step": 25170
+    },
+    {
+      "epoch": 0.5930008007159342,
+      "grad_norm": 2.0071730613708496,
+      "learning_rate": 8.144694079412181e-05,
+      "loss": 2.0596,
+      "step": 25180
+    },
+    {
+      "epoch": 0.5932363054024775,
+      "grad_norm": 1.9122415781021118,
+      "learning_rate": 8.139983985681315e-05,
+      "loss": 1.9869,
+      "step": 25190
+    },
+    {
+      "epoch": 0.5934718100890207,
+      "grad_norm": 2.8985416889190674,
+      "learning_rate": 8.13527389195045e-05,
+      "loss": 1.9928,
+      "step": 25200
+    },
+    {
+      "epoch": 0.593707314775564,
+      "grad_norm": 2.471587657928467,
+      "learning_rate": 8.130563798219585e-05,
+      "loss": 1.9349,
+      "step": 25210
+    },
+    {
+      "epoch": 0.5939428194621073,
+      "grad_norm": 2.232741594314575,
+      "learning_rate": 8.12585370448872e-05,
+      "loss": 2.3072,
+      "step": 25220
+    },
+    {
+      "epoch": 0.5941783241486506,
+      "grad_norm": 2.7327969074249268,
+      "learning_rate": 8.121143610757854e-05,
+      "loss": 2.0982,
+      "step": 25230
+    },
+    {
+      "epoch": 0.5944138288351938,
+      "grad_norm": 2.561145305633545,
+      "learning_rate": 8.116433517026989e-05,
+      "loss": 1.9554,
+      "step": 25240
+    },
+    {
+      "epoch": 0.5946493335217371,
+      "grad_norm": 2.3863377571105957,
+      "learning_rate": 8.111723423296124e-05,
+      "loss": 1.9331,
+      "step": 25250
+    },
+    {
+      "epoch": 0.5948848382082803,
+      "grad_norm": 2.0999460220336914,
+      "learning_rate": 8.107484338938345e-05,
+      "loss": 1.9116,
+      "step": 25260
+    },
+    {
+      "epoch": 0.5951203428948236,
+      "grad_norm": 2.5434398651123047,
+      "learning_rate": 8.102774245207479e-05,
+      "loss": 2.1479,
+      "step": 25270
+    },
+    {
+      "epoch": 0.5953558475813668,
+      "grad_norm": 2.038438558578491,
+      "learning_rate": 8.098064151476614e-05,
+      "loss": 2.0945,
+      "step": 25280
+    },
+    {
+      "epoch": 0.5955913522679102,
+      "grad_norm": 2.3757026195526123,
+      "learning_rate": 8.09335405774575e-05,
+      "loss": 2.142,
+      "step": 25290
+    },
+    {
+      "epoch": 0.5958268569544534,
+      "grad_norm": 2.268975257873535,
+      "learning_rate": 8.088643964014884e-05,
+      "loss": 2.0171,
+      "step": 25300
+    },
+    {
+      "epoch": 0.5960623616409967,
+      "grad_norm": 2.189600944519043,
+      "learning_rate": 8.083933870284019e-05,
+      "loss": 2.0769,
+      "step": 25310
+    },
+    {
+      "epoch": 0.5962978663275399,
+      "grad_norm": 2.25864577293396,
+      "learning_rate": 8.079223776553154e-05,
+      "loss": 2.0467,
+      "step": 25320
+    },
+    {
+      "epoch": 0.5965333710140832,
+      "grad_norm": 1.8093814849853516,
+      "learning_rate": 8.074513682822289e-05,
+      "loss": 1.9886,
+      "step": 25330
+    },
+    {
+      "epoch": 0.5967688757006264,
+      "grad_norm": 2.4894871711730957,
+      "learning_rate": 8.069803589091424e-05,
+      "loss": 2.0495,
+      "step": 25340
+    },
+    {
+      "epoch": 0.5970043803871697,
+      "grad_norm": 2.965756893157959,
+      "learning_rate": 8.065093495360558e-05,
+      "loss": 2.233,
+      "step": 25350
+    },
+    {
+      "epoch": 0.597239885073713,
+      "grad_norm": 2.278423309326172,
+      "learning_rate": 8.060383401629693e-05,
+      "loss": 2.0152,
+      "step": 25360
+    },
+    {
+      "epoch": 0.5974753897602563,
+      "grad_norm": 2.2974514961242676,
+      "learning_rate": 8.055673307898828e-05,
+      "loss": 2.2189,
+      "step": 25370
+    },
+    {
+      "epoch": 0.5977108944467995,
+      "grad_norm": 2.401662588119507,
+      "learning_rate": 8.050963214167963e-05,
+      "loss": 1.9795,
+      "step": 25380
+    },
+    {
+      "epoch": 0.5979463991333428,
+      "grad_norm": 2.429227113723755,
+      "learning_rate": 8.046253120437097e-05,
+      "loss": 1.9638,
+      "step": 25390
+    },
+    {
+      "epoch": 0.598181903819886,
+      "grad_norm": 2.7841548919677734,
+      "learning_rate": 8.041543026706232e-05,
+      "loss": 2.1462,
+      "step": 25400
+    },
+    {
+      "epoch": 0.5984174085064293,
+      "grad_norm": 2.3505892753601074,
+      "learning_rate": 8.036832932975367e-05,
+      "loss": 2.1676,
+      "step": 25410
+    },
+    {
+      "epoch": 0.5986529131929725,
+      "grad_norm": 1.8642781972885132,
+      "learning_rate": 8.032122839244502e-05,
+      "loss": 2.2374,
+      "step": 25420
+    },
+    {
+      "epoch": 0.5988884178795159,
+      "grad_norm": 2.466585874557495,
+      "learning_rate": 8.027412745513637e-05,
+      "loss": 2.1254,
+      "step": 25430
+    },
+    {
+      "epoch": 0.5991239225660591,
+      "grad_norm": 2.247736692428589,
+      "learning_rate": 8.02270265178277e-05,
+      "loss": 1.8839,
+      "step": 25440
+    },
+    {
+      "epoch": 0.5993594272526023,
+      "grad_norm": 2.2101500034332275,
+      "learning_rate": 8.017992558051905e-05,
+      "loss": 2.0983,
+      "step": 25450
+    },
+    {
+      "epoch": 0.5995949319391456,
+      "grad_norm": 2.6747212409973145,
+      "learning_rate": 8.01328246432104e-05,
+      "loss": 2.0788,
+      "step": 25460
+    },
+    {
+      "epoch": 0.5998304366256888,
+      "grad_norm": 2.289003610610962,
+      "learning_rate": 8.008572370590176e-05,
+      "loss": 2.1053,
+      "step": 25470
+    },
+    {
+      "epoch": 0.6000659413122321,
+      "grad_norm": 2.8795812129974365,
+      "learning_rate": 8.003862276859309e-05,
+      "loss": 1.9574,
+      "step": 25480
+    },
+    {
+      "epoch": 0.6003014459987753,
+      "grad_norm": 1.9603990316390991,
+      "learning_rate": 7.999152183128444e-05,
+      "loss": 2.1414,
+      "step": 25490
+    },
+    {
+      "epoch": 0.6005369506853186,
+      "grad_norm": 1.8929415941238403,
+      "learning_rate": 7.994442089397579e-05,
+      "loss": 2.0685,
+      "step": 25500
+    },
+    {
+      "epoch": 0.6007724553718619,
+      "grad_norm": 2.2865395545959473,
+      "learning_rate": 7.989731995666714e-05,
+      "loss": 1.9385,
+      "step": 25510
+    },
+    {
+      "epoch": 0.6010079600584052,
+      "grad_norm": 1.7141823768615723,
+      "learning_rate": 7.985021901935848e-05,
+      "loss": 1.9145,
+      "step": 25520
+    },
+    {
+      "epoch": 0.6012434647449484,
+      "grad_norm": 2.2416176795959473,
+      "learning_rate": 7.980311808204983e-05,
+      "loss": 1.9598,
+      "step": 25530
+    },
+    {
+      "epoch": 0.6014789694314917,
+      "grad_norm": 1.8684344291687012,
+      "learning_rate": 7.975601714474118e-05,
+      "loss": 2.0327,
+      "step": 25540
+    },
+    {
+      "epoch": 0.6017144741180349,
+      "grad_norm": 2.3493103981018066,
+      "learning_rate": 7.970891620743253e-05,
+      "loss": 1.9675,
+      "step": 25550
+    },
+    {
+      "epoch": 0.6019499788045782,
+      "grad_norm": 2.669125556945801,
+      "learning_rate": 7.966181527012388e-05,
+      "loss": 2.1429,
+      "step": 25560
+    },
+    {
+      "epoch": 0.6021854834911214,
+      "grad_norm": 2.1872849464416504,
+      "learning_rate": 7.961471433281523e-05,
+      "loss": 2.1035,
+      "step": 25570
+    },
+    {
+      "epoch": 0.6024209881776648,
+      "grad_norm": 2.5659477710723877,
+      "learning_rate": 7.956761339550658e-05,
+      "loss": 2.1668,
+      "step": 25580
+    },
+    {
+      "epoch": 0.602656492864208,
+      "grad_norm": 1.815410852432251,
+      "learning_rate": 7.952051245819793e-05,
+      "loss": 1.9685,
+      "step": 25590
+    },
+    {
+      "epoch": 0.6028919975507513,
+      "grad_norm": 2.178764820098877,
+      "learning_rate": 7.947341152088927e-05,
+      "loss": 2.0261,
+      "step": 25600
+    },
+    {
+      "epoch": 0.6031275022372945,
+      "grad_norm": 2.0727059841156006,
+      "learning_rate": 7.942631058358062e-05,
+      "loss": 2.3205,
+      "step": 25610
+    },
+    {
+      "epoch": 0.6033630069238378,
+      "grad_norm": 2.4925308227539062,
+      "learning_rate": 7.937920964627197e-05,
+      "loss": 1.9224,
+      "step": 25620
+    },
+    {
+      "epoch": 0.603598511610381,
+      "grad_norm": 2.7875311374664307,
+      "learning_rate": 7.933210870896332e-05,
+      "loss": 2.1901,
+      "step": 25630
+    },
+    {
+      "epoch": 0.6038340162969243,
+      "grad_norm": 2.135871648788452,
+      "learning_rate": 7.928500777165467e-05,
+      "loss": 2.0171,
+      "step": 25640
+    },
+    {
+      "epoch": 0.6040695209834676,
+      "grad_norm": 2.568267345428467,
+      "learning_rate": 7.9237906834346e-05,
+      "loss": 2.2376,
+      "step": 25650
+    },
+    {
+      "epoch": 0.6043050256700109,
+      "grad_norm": 2.8752918243408203,
+      "learning_rate": 7.919080589703736e-05,
+      "loss": 2.0801,
+      "step": 25660
+    },
+    {
+      "epoch": 0.6045405303565541,
+      "grad_norm": 2.950716257095337,
+      "learning_rate": 7.91437049597287e-05,
+      "loss": 2.1433,
+      "step": 25670
+    },
+    {
+      "epoch": 0.6047760350430974,
+      "grad_norm": 1.806060791015625,
+      "learning_rate": 7.909660402242006e-05,
+      "loss": 2.0425,
+      "step": 25680
+    },
+    {
+      "epoch": 0.6050115397296406,
+      "grad_norm": 2.009774923324585,
+      "learning_rate": 7.904950308511139e-05,
+      "loss": 1.9264,
+      "step": 25690
+    },
+    {
+      "epoch": 0.6052470444161839,
+      "grad_norm": 2.4653844833374023,
+      "learning_rate": 7.900240214780274e-05,
+      "loss": 2.2729,
+      "step": 25700
+    },
+    {
+      "epoch": 0.6054825491027271,
+      "grad_norm": 2.0661404132843018,
+      "learning_rate": 7.895530121049409e-05,
+      "loss": 1.8604,
+      "step": 25710
+    },
+    {
+      "epoch": 0.6057180537892705,
+      "grad_norm": 2.1067216396331787,
+      "learning_rate": 7.890820027318544e-05,
+      "loss": 1.918,
+      "step": 25720
+    },
+    {
+      "epoch": 0.6059535584758137,
+      "grad_norm": 1.9731528759002686,
+      "learning_rate": 7.88610993358768e-05,
+      "loss": 2.0417,
+      "step": 25730
+    },
+    {
+      "epoch": 0.6061890631623569,
+      "grad_norm": 2.0043580532073975,
+      "learning_rate": 7.881399839856813e-05,
+      "loss": 1.8275,
+      "step": 25740
+    },
+    {
+      "epoch": 0.6064245678489002,
+      "grad_norm": 2.3526785373687744,
+      "learning_rate": 7.876689746125948e-05,
+      "loss": 1.9985,
+      "step": 25750
+    },
+    {
+      "epoch": 0.6066600725354434,
+      "grad_norm": 2.168198585510254,
+      "learning_rate": 7.871979652395083e-05,
+      "loss": 1.9785,
+      "step": 25760
+    },
+    {
+      "epoch": 0.6068955772219867,
+      "grad_norm": 1.7501553297042847,
+      "learning_rate": 7.867269558664218e-05,
+      "loss": 2.0646,
+      "step": 25770
+    },
+    {
+      "epoch": 0.6071310819085299,
+      "grad_norm": 2.0135562419891357,
+      "learning_rate": 7.862559464933352e-05,
+      "loss": 2.152,
+      "step": 25780
+    },
+    {
+      "epoch": 0.6073665865950733,
+      "grad_norm": 1.9424461126327515,
+      "learning_rate": 7.857849371202487e-05,
+      "loss": 2.0319,
+      "step": 25790
+    },
+    {
+      "epoch": 0.6076020912816165,
+      "grad_norm": 2.0011091232299805,
+      "learning_rate": 7.853139277471622e-05,
+      "loss": 1.9419,
+      "step": 25800
+    },
+    {
+      "epoch": 0.6078375959681598,
+      "grad_norm": 1.9479938745498657,
+      "learning_rate": 7.848429183740757e-05,
+      "loss": 2.0519,
+      "step": 25810
+    },
+    {
+      "epoch": 0.608073100654703,
+      "grad_norm": 2.629399538040161,
+      "learning_rate": 7.84371909000989e-05,
+      "loss": 1.8832,
+      "step": 25820
+    },
+    {
+      "epoch": 0.6083086053412463,
+      "grad_norm": 2.2955732345581055,
+      "learning_rate": 7.839008996279026e-05,
+      "loss": 1.8042,
+      "step": 25830
+    },
+    {
+      "epoch": 0.6085441100277895,
+      "grad_norm": 2.8424713611602783,
+      "learning_rate": 7.83429890254816e-05,
+      "loss": 2.4232,
+      "step": 25840
+    },
+    {
+      "epoch": 0.6087796147143328,
+      "grad_norm": 2.084066152572632,
+      "learning_rate": 7.829588808817296e-05,
+      "loss": 2.0076,
+      "step": 25850
+    },
+    {
+      "epoch": 0.609015119400876,
+      "grad_norm": 2.988767623901367,
+      "learning_rate": 7.82487871508643e-05,
+      "loss": 1.9684,
+      "step": 25860
+    },
+    {
+      "epoch": 0.6092506240874194,
+      "grad_norm": 2.3572447299957275,
+      "learning_rate": 7.820168621355566e-05,
+      "loss": 1.9121,
+      "step": 25870
+    },
+    {
+      "epoch": 0.6094861287739626,
+      "grad_norm": 1.948307752609253,
+      "learning_rate": 7.8154585276247e-05,
+      "loss": 1.6575,
+      "step": 25880
+    },
+    {
+      "epoch": 0.6097216334605059,
+      "grad_norm": 4.080255031585693,
+      "learning_rate": 7.810748433893836e-05,
+      "loss": 2.0366,
+      "step": 25890
+    },
+    {
+      "epoch": 0.6099571381470491,
+      "grad_norm": 2.5627658367156982,
+      "learning_rate": 7.80603834016297e-05,
+      "loss": 1.7901,
+      "step": 25900
+    },
+    {
+      "epoch": 0.6101926428335924,
+      "grad_norm": 1.9693503379821777,
+      "learning_rate": 7.801328246432104e-05,
+      "loss": 1.9767,
+      "step": 25910
+    },
+    {
+      "epoch": 0.6104281475201356,
+      "grad_norm": 2.0714879035949707,
+      "learning_rate": 7.79661815270124e-05,
+      "loss": 2.0898,
+      "step": 25920
+    },
+    {
+      "epoch": 0.610663652206679,
+      "grad_norm": 2.8542380332946777,
+      "learning_rate": 7.791908058970374e-05,
+      "loss": 2.2137,
+      "step": 25930
+    },
+    {
+      "epoch": 0.6108991568932222,
+      "grad_norm": 2.3239822387695312,
+      "learning_rate": 7.78719796523951e-05,
+      "loss": 1.8162,
+      "step": 25940
+    },
+    {
+      "epoch": 0.6111346615797655,
+      "grad_norm": 1.9721574783325195,
+      "learning_rate": 7.782487871508643e-05,
+      "loss": 2.1246,
+      "step": 25950
+    },
+    {
+      "epoch": 0.6113701662663087,
+      "grad_norm": 2.5218284130096436,
+      "learning_rate": 7.777777777777778e-05,
+      "loss": 1.9872,
+      "step": 25960
+    },
+    {
+      "epoch": 0.611605670952852,
+      "grad_norm": 1.9830020666122437,
+      "learning_rate": 7.773067684046913e-05,
+      "loss": 1.8981,
+      "step": 25970
+    },
+    {
+      "epoch": 0.6118411756393952,
+      "grad_norm": 2.9052417278289795,
+      "learning_rate": 7.768357590316048e-05,
+      "loss": 1.961,
+      "step": 25980
+    },
+    {
+      "epoch": 0.6120766803259385,
+      "grad_norm": 2.380767583847046,
+      "learning_rate": 7.763647496585182e-05,
+      "loss": 2.0117,
+      "step": 25990
+    },
+    {
+      "epoch": 0.6123121850124817,
+      "grad_norm": 2.563429117202759,
+      "learning_rate": 7.758937402854317e-05,
+      "loss": 2.0541,
+      "step": 26000
+    },
+    {
+      "epoch": 0.612547689699025,
+      "grad_norm": 2.067661762237549,
+      "learning_rate": 7.754227309123452e-05,
+      "loss": 1.9734,
+      "step": 26010
+    },
+    {
+      "epoch": 0.6127831943855683,
+      "grad_norm": 2.019895315170288,
+      "learning_rate": 7.749517215392587e-05,
+      "loss": 2.2234,
+      "step": 26020
+    },
+    {
+      "epoch": 0.6130186990721115,
+      "grad_norm": 2.9683279991149902,
+      "learning_rate": 7.74480712166172e-05,
+      "loss": 1.9863,
+      "step": 26030
+    },
+    {
+      "epoch": 0.6132542037586548,
+      "grad_norm": 2.322253465652466,
+      "learning_rate": 7.740097027930856e-05,
+      "loss": 2.0671,
+      "step": 26040
+    },
+    {
+      "epoch": 0.613489708445198,
+      "grad_norm": 2.235363483428955,
+      "learning_rate": 7.73538693419999e-05,
+      "loss": 1.9467,
+      "step": 26050
+    },
+    {
+      "epoch": 0.6137252131317413,
+      "grad_norm": 2.408618927001953,
+      "learning_rate": 7.730676840469126e-05,
+      "loss": 2.1104,
+      "step": 26060
+    },
+    {
+      "epoch": 0.6139607178182845,
+      "grad_norm": 2.392561912536621,
+      "learning_rate": 7.725966746738261e-05,
+      "loss": 1.8468,
+      "step": 26070
+    },
+    {
+      "epoch": 0.6141962225048279,
+      "grad_norm": 2.707282781600952,
+      "learning_rate": 7.721256653007394e-05,
+      "loss": 2.0011,
+      "step": 26080
+    },
+    {
+      "epoch": 0.6144317271913711,
+      "grad_norm": 2.337557792663574,
+      "learning_rate": 7.71654655927653e-05,
+      "loss": 2.026,
+      "step": 26090
+    },
+    {
+      "epoch": 0.6146672318779144,
+      "grad_norm": 2.9671690464019775,
+      "learning_rate": 7.711836465545664e-05,
+      "loss": 1.965,
+      "step": 26100
+    },
+    {
+      "epoch": 0.6149027365644576,
+      "grad_norm": 1.8119666576385498,
+      "learning_rate": 7.7071263718148e-05,
+      "loss": 2.0404,
+      "step": 26110
+    },
+    {
+      "epoch": 0.6151382412510009,
+      "grad_norm": 2.940488576889038,
+      "learning_rate": 7.702416278083933e-05,
+      "loss": 1.7759,
+      "step": 26120
+    },
+    {
+      "epoch": 0.6153737459375441,
+      "grad_norm": 2.25976300239563,
+      "learning_rate": 7.697706184353068e-05,
+      "loss": 1.7815,
+      "step": 26130
+    },
+    {
+      "epoch": 0.6156092506240874,
+      "grad_norm": 2.183560371398926,
+      "learning_rate": 7.692996090622205e-05,
+      "loss": 1.8272,
+      "step": 26140
+    },
+    {
+      "epoch": 0.6158447553106307,
+      "grad_norm": 2.320969820022583,
+      "learning_rate": 7.68828599689134e-05,
+      "loss": 2.1829,
+      "step": 26150
+    },
+    {
+      "epoch": 0.616080259997174,
+      "grad_norm": 2.730713129043579,
+      "learning_rate": 7.683575903160473e-05,
+      "loss": 1.908,
+      "step": 26160
+    },
+    {
+      "epoch": 0.6163157646837172,
+      "grad_norm": 2.4461238384246826,
+      "learning_rate": 7.678865809429608e-05,
+      "loss": 2.1768,
+      "step": 26170
+    },
+    {
+      "epoch": 0.6165512693702605,
+      "grad_norm": 2.157590866088867,
+      "learning_rate": 7.674155715698743e-05,
+      "loss": 2.1194,
+      "step": 26180
+    },
+    {
+      "epoch": 0.6167867740568037,
+      "grad_norm": 2.2713019847869873,
+      "learning_rate": 7.669445621967878e-05,
+      "loss": 2.0888,
+      "step": 26190
+    },
+    {
+      "epoch": 0.617022278743347,
+      "grad_norm": 2.798720598220825,
+      "learning_rate": 7.664735528237012e-05,
+      "loss": 2.0496,
+      "step": 26200
+    },
+    {
+      "epoch": 0.6172577834298902,
+      "grad_norm": 2.4818081855773926,
+      "learning_rate": 7.660025434506147e-05,
+      "loss": 2.2425,
+      "step": 26210
+    },
+    {
+      "epoch": 0.6174932881164336,
+      "grad_norm": 2.5907058715820312,
+      "learning_rate": 7.655315340775282e-05,
+      "loss": 1.9915,
+      "step": 26220
+    },
+    {
+      "epoch": 0.6177287928029768,
+      "grad_norm": 3.659034490585327,
+      "learning_rate": 7.650605247044417e-05,
+      "loss": 1.9745,
+      "step": 26230
+    },
+    {
+      "epoch": 0.6179642974895201,
+      "grad_norm": 1.8911350965499878,
+      "learning_rate": 7.645895153313552e-05,
+      "loss": 2.02,
+      "step": 26240
+    },
+    {
+      "epoch": 0.6181998021760633,
+      "grad_norm": 2.086848020553589,
+      "learning_rate": 7.641185059582686e-05,
+      "loss": 2.0886,
+      "step": 26250
+    },
+    {
+      "epoch": 0.6184353068626066,
+      "grad_norm": 2.335862636566162,
+      "learning_rate": 7.636474965851821e-05,
+      "loss": 2.13,
+      "step": 26260
+    },
+    {
+      "epoch": 0.6186708115491498,
+      "grad_norm": 3.3991475105285645,
+      "learning_rate": 7.631764872120956e-05,
+      "loss": 2.0224,
+      "step": 26270
+    },
+    {
+      "epoch": 0.6189063162356931,
+      "grad_norm": 2.554853677749634,
+      "learning_rate": 7.627054778390091e-05,
+      "loss": 2.0263,
+      "step": 26280
+    },
+    {
+      "epoch": 0.6191418209222364,
+      "grad_norm": 1.8681471347808838,
+      "learning_rate": 7.622344684659224e-05,
+      "loss": 1.8256,
+      "step": 26290
+    },
+    {
+      "epoch": 0.6193773256087796,
+      "grad_norm": 2.5911004543304443,
+      "learning_rate": 7.61763459092836e-05,
+      "loss": 2.1467,
+      "step": 26300
+    },
+    {
+      "epoch": 0.6196128302953229,
+      "grad_norm": 2.3044159412384033,
+      "learning_rate": 7.612924497197495e-05,
+      "loss": 1.8694,
+      "step": 26310
+    },
+    {
+      "epoch": 0.6198483349818661,
+      "grad_norm": 2.6165413856506348,
+      "learning_rate": 7.60821440346663e-05,
+      "loss": 1.9561,
+      "step": 26320
+    },
+    {
+      "epoch": 0.6200838396684094,
+      "grad_norm": 2.0098507404327393,
+      "learning_rate": 7.603504309735763e-05,
+      "loss": 2.0378,
+      "step": 26330
+    },
+    {
+      "epoch": 0.6203193443549526,
+      "grad_norm": 3.309293508529663,
+      "learning_rate": 7.598794216004898e-05,
+      "loss": 1.8842,
+      "step": 26340
+    },
+    {
+      "epoch": 0.6205548490414959,
+      "grad_norm": 2.2279319763183594,
+      "learning_rate": 7.594084122274033e-05,
+      "loss": 2.1463,
+      "step": 26350
+    },
+    {
+      "epoch": 0.6207903537280391,
+      "grad_norm": 1.7420648336410522,
+      "learning_rate": 7.589374028543168e-05,
+      "loss": 2.033,
+      "step": 26360
+    },
+    {
+      "epoch": 0.6210258584145825,
+      "grad_norm": 2.068329334259033,
+      "learning_rate": 7.584663934812303e-05,
+      "loss": 2.0667,
+      "step": 26370
+    },
+    {
+      "epoch": 0.6212613631011257,
+      "grad_norm": 2.148442506790161,
+      "learning_rate": 7.579953841081437e-05,
+      "loss": 2.0649,
+      "step": 26380
+    },
+    {
+      "epoch": 0.621496867787669,
+      "grad_norm": 2.486476421356201,
+      "learning_rate": 7.575243747350572e-05,
+      "loss": 2.0631,
+      "step": 26390
+    },
+    {
+      "epoch": 0.6217323724742122,
+      "grad_norm": 2.4887843132019043,
+      "learning_rate": 7.570533653619707e-05,
+      "loss": 1.7455,
+      "step": 26400
+    },
+    {
+      "epoch": 0.6219678771607555,
+      "grad_norm": 1.8628402948379517,
+      "learning_rate": 7.565823559888842e-05,
+      "loss": 2.0665,
+      "step": 26410
+    },
+    {
+      "epoch": 0.6222033818472987,
+      "grad_norm": 1.9637489318847656,
+      "learning_rate": 7.561113466157977e-05,
+      "loss": 1.9645,
+      "step": 26420
+    },
+    {
+      "epoch": 0.622438886533842,
+      "grad_norm": 2.734393835067749,
+      "learning_rate": 7.556403372427112e-05,
+      "loss": 2.0079,
+      "step": 26430
+    },
+    {
+      "epoch": 0.6226743912203853,
+      "grad_norm": 2.7208571434020996,
+      "learning_rate": 7.551693278696247e-05,
+      "loss": 2.0066,
+      "step": 26440
+    },
+    {
+      "epoch": 0.6229098959069286,
+      "grad_norm": 2.311509132385254,
+      "learning_rate": 7.546983184965382e-05,
+      "loss": 2.0782,
+      "step": 26450
+    },
+    {
+      "epoch": 0.6231454005934718,
+      "grad_norm": 2.43361759185791,
+      "learning_rate": 7.542273091234516e-05,
+      "loss": 2.1844,
+      "step": 26460
+    },
+    {
+      "epoch": 0.6233809052800151,
+      "grad_norm": 2.6297848224639893,
+      "learning_rate": 7.537562997503651e-05,
+      "loss": 1.9422,
+      "step": 26470
+    },
+    {
+      "epoch": 0.6236164099665583,
+      "grad_norm": 1.997184157371521,
+      "learning_rate": 7.532852903772786e-05,
+      "loss": 2.1683,
+      "step": 26480
+    },
+    {
+      "epoch": 0.6238519146531016,
+      "grad_norm": 1.8917272090911865,
+      "learning_rate": 7.528142810041921e-05,
+      "loss": 1.9385,
+      "step": 26490
+    },
+    {
+      "epoch": 0.6240874193396448,
+      "grad_norm": 2.0081610679626465,
+      "learning_rate": 7.523432716311055e-05,
+      "loss": 1.9071,
+      "step": 26500
+    },
+    {
+      "epoch": 0.6243229240261882,
+      "grad_norm": 2.179267644882202,
+      "learning_rate": 7.51872262258019e-05,
+      "loss": 2.1131,
+      "step": 26510
+    },
+    {
+      "epoch": 0.6245584287127314,
+      "grad_norm": 1.9860879182815552,
+      "learning_rate": 7.514012528849325e-05,
+      "loss": 2.3245,
+      "step": 26520
+    },
+    {
+      "epoch": 0.6247939333992747,
+      "grad_norm": 1.9527974128723145,
+      "learning_rate": 7.50930243511846e-05,
+      "loss": 1.9865,
+      "step": 26530
+    },
+    {
+      "epoch": 0.6250294380858179,
+      "grad_norm": 2.507073402404785,
+      "learning_rate": 7.504592341387595e-05,
+      "loss": 1.8888,
+      "step": 26540
+    },
+    {
+      "epoch": 0.6252649427723612,
+      "grad_norm": 2.8235909938812256,
+      "learning_rate": 7.499882247656728e-05,
+      "loss": 2.0307,
+      "step": 26550
+    },
+    {
+      "epoch": 0.6255004474589044,
+      "grad_norm": 3.5919525623321533,
+      "learning_rate": 7.495172153925863e-05,
+      "loss": 2.1907,
+      "step": 26560
+    },
+    {
+      "epoch": 0.6257359521454477,
+      "grad_norm": 1.6083260774612427,
+      "learning_rate": 7.490462060194998e-05,
+      "loss": 1.8715,
+      "step": 26570
+    },
+    {
+      "epoch": 0.625971456831991,
+      "grad_norm": 2.9224612712860107,
+      "learning_rate": 7.485751966464133e-05,
+      "loss": 2.1417,
+      "step": 26580
+    },
+    {
+      "epoch": 0.6262069615185342,
+      "grad_norm": 2.0312814712524414,
+      "learning_rate": 7.481041872733267e-05,
+      "loss": 1.9067,
+      "step": 26590
+    },
+    {
+      "epoch": 0.6264424662050775,
+      "grad_norm": 2.3932294845581055,
+      "learning_rate": 7.476331779002402e-05,
+      "loss": 2.0652,
+      "step": 26600
+    },
+    {
+      "epoch": 0.6266779708916207,
+      "grad_norm": 2.058734893798828,
+      "learning_rate": 7.471621685271537e-05,
+      "loss": 2.1212,
+      "step": 26610
+    },
+    {
+      "epoch": 0.626913475578164,
+      "grad_norm": 2.9603049755096436,
+      "learning_rate": 7.466911591540672e-05,
+      "loss": 1.917,
+      "step": 26620
+    },
+    {
+      "epoch": 0.6271489802647072,
+      "grad_norm": 2.741225004196167,
+      "learning_rate": 7.462201497809806e-05,
+      "loss": 2.095,
+      "step": 26630
+    },
+    {
+      "epoch": 0.6273844849512505,
+      "grad_norm": 2.8252007961273193,
+      "learning_rate": 7.457491404078941e-05,
+      "loss": 2.045,
+      "step": 26640
+    },
+    {
+      "epoch": 0.6276199896377938,
+      "grad_norm": 2.179786443710327,
+      "learning_rate": 7.452781310348076e-05,
+      "loss": 2.0414,
+      "step": 26650
+    },
+    {
+      "epoch": 0.6278554943243371,
+      "grad_norm": 2.8023617267608643,
+      "learning_rate": 7.448071216617211e-05,
+      "loss": 2.2393,
+      "step": 26660
+    },
+    {
+      "epoch": 0.6280909990108803,
+      "grad_norm": 1.851933240890503,
+      "learning_rate": 7.443361122886346e-05,
+      "loss": 2.1044,
+      "step": 26670
+    },
+    {
+      "epoch": 0.6283265036974236,
+      "grad_norm": 2.2054028511047363,
+      "learning_rate": 7.43865102915548e-05,
+      "loss": 2.328,
+      "step": 26680
+    },
+    {
+      "epoch": 0.6285620083839668,
+      "grad_norm": 2.3730199337005615,
+      "learning_rate": 7.433940935424615e-05,
+      "loss": 1.9667,
+      "step": 26690
+    },
+    {
+      "epoch": 0.6287975130705101,
+      "grad_norm": 2.3404037952423096,
+      "learning_rate": 7.429230841693751e-05,
+      "loss": 2.0277,
+      "step": 26700
+    },
+    {
+      "epoch": 0.6290330177570533,
+      "grad_norm": 2.6131813526153564,
+      "learning_rate": 7.424520747962885e-05,
+      "loss": 2.0922,
+      "step": 26710
+    },
+    {
+      "epoch": 0.6292685224435967,
+      "grad_norm": 1.9105520248413086,
+      "learning_rate": 7.41981065423202e-05,
+      "loss": 2.2015,
+      "step": 26720
+    },
+    {
+      "epoch": 0.6295040271301399,
+      "grad_norm": 2.223731517791748,
+      "learning_rate": 7.415100560501155e-05,
+      "loss": 1.8634,
+      "step": 26730
+    },
+    {
+      "epoch": 0.6297395318166832,
+      "grad_norm": 2.680253028869629,
+      "learning_rate": 7.41039046677029e-05,
+      "loss": 2.0635,
+      "step": 26740
+    },
+    {
+      "epoch": 0.6299750365032264,
+      "grad_norm": 3.308995008468628,
+      "learning_rate": 7.405680373039425e-05,
+      "loss": 1.8926,
+      "step": 26750
+    },
+    {
+      "epoch": 0.6302105411897697,
+      "grad_norm": 2.192030668258667,
+      "learning_rate": 7.400970279308558e-05,
+      "loss": 1.7986,
+      "step": 26760
+    },
+    {
+      "epoch": 0.6304460458763129,
+      "grad_norm": 2.320146322250366,
+      "learning_rate": 7.396260185577693e-05,
+      "loss": 2.1723,
+      "step": 26770
+    },
+    {
+      "epoch": 0.6306815505628562,
+      "grad_norm": 2.5888264179229736,
+      "learning_rate": 7.391550091846828e-05,
+      "loss": 1.9864,
+      "step": 26780
+    },
+    {
+      "epoch": 0.6309170552493994,
+      "grad_norm": 2.5209715366363525,
+      "learning_rate": 7.386839998115963e-05,
+      "loss": 2.0133,
+      "step": 26790
+    },
+    {
+      "epoch": 0.6311525599359428,
+      "grad_norm": 2.5114974975585938,
+      "learning_rate": 7.382129904385097e-05,
+      "loss": 2.0248,
+      "step": 26800
+    },
+    {
+      "epoch": 0.631388064622486,
+      "grad_norm": 1.4139314889907837,
+      "learning_rate": 7.377419810654232e-05,
+      "loss": 1.7762,
+      "step": 26810
+    },
+    {
+      "epoch": 0.6316235693090293,
+      "grad_norm": 2.1081104278564453,
+      "learning_rate": 7.372709716923367e-05,
+      "loss": 1.9416,
+      "step": 26820
+    },
+    {
+      "epoch": 0.6318590739955725,
+      "grad_norm": 2.1463840007781982,
+      "learning_rate": 7.367999623192502e-05,
+      "loss": 1.8741,
+      "step": 26830
+    },
+    {
+      "epoch": 0.6320945786821158,
+      "grad_norm": 2.403310775756836,
+      "learning_rate": 7.363289529461637e-05,
+      "loss": 1.974,
+      "step": 26840
+    },
+    {
+      "epoch": 0.632330083368659,
+      "grad_norm": 2.203840732574463,
+      "learning_rate": 7.358579435730771e-05,
+      "loss": 1.9,
+      "step": 26850
+    },
+    {
+      "epoch": 0.6325655880552022,
+      "grad_norm": 2.0691256523132324,
+      "learning_rate": 7.353869341999906e-05,
+      "loss": 1.9828,
+      "step": 26860
+    },
+    {
+      "epoch": 0.6328010927417456,
+      "grad_norm": 1.5303398370742798,
+      "learning_rate": 7.349159248269041e-05,
+      "loss": 1.8315,
+      "step": 26870
+    },
+    {
+      "epoch": 0.6330365974282888,
+      "grad_norm": 2.035423994064331,
+      "learning_rate": 7.344449154538176e-05,
+      "loss": 2.2133,
+      "step": 26880
+    },
+    {
+      "epoch": 0.6332721021148321,
+      "grad_norm": 2.0543622970581055,
+      "learning_rate": 7.33973906080731e-05,
+      "loss": 1.8439,
+      "step": 26890
+    },
+    {
+      "epoch": 0.6335076068013753,
+      "grad_norm": 2.3338544368743896,
+      "learning_rate": 7.335028967076445e-05,
+      "loss": 2.1695,
+      "step": 26900
+    },
+    {
+      "epoch": 0.6337431114879186,
+      "grad_norm": 2.5324740409851074,
+      "learning_rate": 7.33031887334558e-05,
+      "loss": 2.1254,
+      "step": 26910
+    },
+    {
+      "epoch": 0.6339786161744618,
+      "grad_norm": 2.218935489654541,
+      "learning_rate": 7.325608779614715e-05,
+      "loss": 2.0474,
+      "step": 26920
+    },
+    {
+      "epoch": 0.6342141208610051,
+      "grad_norm": 2.561553478240967,
+      "learning_rate": 7.320898685883848e-05,
+      "loss": 2.3904,
+      "step": 26930
+    },
+    {
+      "epoch": 0.6344496255475484,
+      "grad_norm": 2.866525411605835,
+      "learning_rate": 7.316188592152983e-05,
+      "loss": 2.0126,
+      "step": 26940
+    },
+    {
+      "epoch": 0.6346851302340917,
+      "grad_norm": 1.68939208984375,
+      "learning_rate": 7.311478498422118e-05,
+      "loss": 2.0345,
+      "step": 26950
+    },
+    {
+      "epoch": 0.6349206349206349,
+      "grad_norm": 2.2209982872009277,
+      "learning_rate": 7.306768404691253e-05,
+      "loss": 2.2129,
+      "step": 26960
+    },
+    {
+      "epoch": 0.6351561396071782,
+      "grad_norm": 2.152632713317871,
+      "learning_rate": 7.302058310960389e-05,
+      "loss": 2.0254,
+      "step": 26970
+    },
+    {
+      "epoch": 0.6353916442937214,
+      "grad_norm": 2.0267491340637207,
+      "learning_rate": 7.297348217229522e-05,
+      "loss": 2.1108,
+      "step": 26980
+    },
+    {
+      "epoch": 0.6356271489802647,
+      "grad_norm": 1.9867489337921143,
+      "learning_rate": 7.292638123498659e-05,
+      "loss": 2.1763,
+      "step": 26990
+    },
+    {
+      "epoch": 0.6358626536668079,
+      "grad_norm": 3.316333532333374,
+      "learning_rate": 7.287928029767794e-05,
+      "loss": 2.1017,
+      "step": 27000
+    },
+    {
+      "epoch": 0.6360981583533513,
+      "grad_norm": 1.8042023181915283,
+      "learning_rate": 7.283217936036927e-05,
+      "loss": 2.1005,
+      "step": 27010
+    },
+    {
+      "epoch": 0.6363336630398945,
+      "grad_norm": 2.227996587753296,
+      "learning_rate": 7.278507842306062e-05,
+      "loss": 2.1378,
+      "step": 27020
+    },
+    {
+      "epoch": 0.6365691677264378,
+      "grad_norm": 2.519883155822754,
+      "learning_rate": 7.273797748575197e-05,
+      "loss": 1.8913,
+      "step": 27030
+    },
+    {
+      "epoch": 0.636804672412981,
+      "grad_norm": 1.6812520027160645,
+      "learning_rate": 7.269087654844332e-05,
+      "loss": 1.8019,
+      "step": 27040
+    },
+    {
+      "epoch": 0.6370401770995243,
+      "grad_norm": 3.9242050647735596,
+      "learning_rate": 7.264377561113467e-05,
+      "loss": 2.3015,
+      "step": 27050
+    },
+    {
+      "epoch": 0.6372756817860675,
+      "grad_norm": 1.8133604526519775,
+      "learning_rate": 7.259667467382601e-05,
+      "loss": 1.9485,
+      "step": 27060
+    },
+    {
+      "epoch": 0.6375111864726108,
+      "grad_norm": 2.5527639389038086,
+      "learning_rate": 7.254957373651736e-05,
+      "loss": 2.0823,
+      "step": 27070
+    },
+    {
+      "epoch": 0.637746691159154,
+      "grad_norm": 2.00186824798584,
+      "learning_rate": 7.250247279920871e-05,
+      "loss": 2.0701,
+      "step": 27080
+    },
+    {
+      "epoch": 0.6379821958456974,
+      "grad_norm": 2.2295942306518555,
+      "learning_rate": 7.245537186190006e-05,
+      "loss": 2.0213,
+      "step": 27090
+    },
+    {
+      "epoch": 0.6382177005322406,
+      "grad_norm": 2.3956990242004395,
+      "learning_rate": 7.24082709245914e-05,
+      "loss": 1.8728,
+      "step": 27100
+    },
+    {
+      "epoch": 0.6384532052187839,
+      "grad_norm": 2.1132164001464844,
+      "learning_rate": 7.236116998728275e-05,
+      "loss": 1.8988,
+      "step": 27110
+    },
+    {
+      "epoch": 0.6386887099053271,
+      "grad_norm": 2.5249156951904297,
+      "learning_rate": 7.23140690499741e-05,
+      "loss": 2.051,
+      "step": 27120
+    },
+    {
+      "epoch": 0.6389242145918704,
+      "grad_norm": 2.1515731811523438,
+      "learning_rate": 7.226696811266545e-05,
+      "loss": 1.8113,
+      "step": 27130
+    },
+    {
+      "epoch": 0.6391597192784136,
+      "grad_norm": 1.9574979543685913,
+      "learning_rate": 7.22198671753568e-05,
+      "loss": 2.2399,
+      "step": 27140
+    },
+    {
+      "epoch": 0.6393952239649568,
+      "grad_norm": 2.0978519916534424,
+      "learning_rate": 7.217276623804814e-05,
+      "loss": 2.333,
+      "step": 27150
+    },
+    {
+      "epoch": 0.6396307286515002,
+      "grad_norm": 1.8691319227218628,
+      "learning_rate": 7.212566530073949e-05,
+      "loss": 2.0774,
+      "step": 27160
+    },
+    {
+      "epoch": 0.6398662333380434,
+      "grad_norm": 1.9516493082046509,
+      "learning_rate": 7.207856436343084e-05,
+      "loss": 2.2324,
+      "step": 27170
+    },
+    {
+      "epoch": 0.6401017380245867,
+      "grad_norm": 1.61760675907135,
+      "learning_rate": 7.203146342612219e-05,
+      "loss": 2.127,
+      "step": 27180
+    },
+    {
+      "epoch": 0.6403372427111299,
+      "grad_norm": 1.9777883291244507,
+      "learning_rate": 7.198436248881352e-05,
+      "loss": 2.0067,
+      "step": 27190
+    },
+    {
+      "epoch": 0.6405727473976732,
+      "grad_norm": 2.270724058151245,
+      "learning_rate": 7.193726155150487e-05,
+      "loss": 1.8146,
+      "step": 27200
+    },
+    {
+      "epoch": 0.6408082520842164,
+      "grad_norm": 2.2753586769104004,
+      "learning_rate": 7.189016061419622e-05,
+      "loss": 2.1406,
+      "step": 27210
+    },
+    {
+      "epoch": 0.6410437567707598,
+      "grad_norm": 2.114849328994751,
+      "learning_rate": 7.184305967688757e-05,
+      "loss": 1.9977,
+      "step": 27220
+    },
+    {
+      "epoch": 0.641279261457303,
+      "grad_norm": 2.480898857116699,
+      "learning_rate": 7.179595873957891e-05,
+      "loss": 1.8545,
+      "step": 27230
+    },
+    {
+      "epoch": 0.6415147661438463,
+      "grad_norm": 2.312798500061035,
+      "learning_rate": 7.174885780227026e-05,
+      "loss": 2.0963,
+      "step": 27240
+    },
+    {
+      "epoch": 0.6417502708303895,
+      "grad_norm": 2.0719785690307617,
+      "learning_rate": 7.170175686496161e-05,
+      "loss": 2.0581,
+      "step": 27250
+    },
+    {
+      "epoch": 0.6419857755169328,
+      "grad_norm": 2.2538905143737793,
+      "learning_rate": 7.165465592765296e-05,
+      "loss": 1.9055,
+      "step": 27260
+    },
+    {
+      "epoch": 0.642221280203476,
+      "grad_norm": 2.6148452758789062,
+      "learning_rate": 7.160755499034431e-05,
+      "loss": 2.2561,
+      "step": 27270
+    },
+    {
+      "epoch": 0.6424567848900193,
+      "grad_norm": 1.9259710311889648,
+      "learning_rate": 7.156045405303566e-05,
+      "loss": 2.2007,
+      "step": 27280
+    },
+    {
+      "epoch": 0.6426922895765625,
+      "grad_norm": 1.789009690284729,
+      "learning_rate": 7.151335311572701e-05,
+      "loss": 1.9831,
+      "step": 27290
+    },
+    {
+      "epoch": 0.6429277942631059,
+      "grad_norm": 3.458051919937134,
+      "learning_rate": 7.146625217841836e-05,
+      "loss": 2.0307,
+      "step": 27300
+    },
+    {
+      "epoch": 0.6431632989496491,
+      "grad_norm": 2.757291078567505,
+      "learning_rate": 7.14191512411097e-05,
+      "loss": 2.1219,
+      "step": 27310
+    },
+    {
+      "epoch": 0.6433988036361924,
+      "grad_norm": 1.8680845499038696,
+      "learning_rate": 7.137205030380105e-05,
+      "loss": 2.0296,
+      "step": 27320
+    },
+    {
+      "epoch": 0.6436343083227356,
+      "grad_norm": 2.32859468460083,
+      "learning_rate": 7.13249493664924e-05,
+      "loss": 2.1007,
+      "step": 27330
+    },
+    {
+      "epoch": 0.6438698130092789,
+      "grad_norm": 2.07627272605896,
+      "learning_rate": 7.127784842918375e-05,
+      "loss": 1.9358,
+      "step": 27340
+    },
+    {
+      "epoch": 0.6441053176958221,
+      "grad_norm": 2.0837743282318115,
+      "learning_rate": 7.12307474918751e-05,
+      "loss": 1.8441,
+      "step": 27350
+    },
+    {
+      "epoch": 0.6443408223823655,
+      "grad_norm": 2.417036771774292,
+      "learning_rate": 7.118364655456644e-05,
+      "loss": 1.9943,
+      "step": 27360
+    },
+    {
+      "epoch": 0.6445763270689087,
+      "grad_norm": 2.3873021602630615,
+      "learning_rate": 7.113654561725779e-05,
+      "loss": 2.0055,
+      "step": 27370
+    },
+    {
+      "epoch": 0.644811831755452,
+      "grad_norm": 1.9441145658493042,
+      "learning_rate": 7.108944467994914e-05,
+      "loss": 2.0131,
+      "step": 27380
+    },
+    {
+      "epoch": 0.6450473364419952,
+      "grad_norm": 2.0448830127716064,
+      "learning_rate": 7.104234374264049e-05,
+      "loss": 2.2472,
+      "step": 27390
+    },
+    {
+      "epoch": 0.6452828411285385,
+      "grad_norm": 2.079862594604492,
+      "learning_rate": 7.099524280533182e-05,
+      "loss": 1.8943,
+      "step": 27400
+    },
+    {
+      "epoch": 0.6455183458150817,
+      "grad_norm": 3.530986785888672,
+      "learning_rate": 7.094814186802317e-05,
+      "loss": 2.0522,
+      "step": 27410
+    },
+    {
+      "epoch": 0.645753850501625,
+      "grad_norm": 2.296212673187256,
+      "learning_rate": 7.090104093071452e-05,
+      "loss": 1.7455,
+      "step": 27420
+    },
+    {
+      "epoch": 0.6459893551881682,
+      "grad_norm": 2.3414154052734375,
+      "learning_rate": 7.085393999340587e-05,
+      "loss": 2.0542,
+      "step": 27430
+    },
+    {
+      "epoch": 0.6462248598747115,
+      "grad_norm": 2.2467596530914307,
+      "learning_rate": 7.080683905609722e-05,
+      "loss": 2.0408,
+      "step": 27440
+    },
+    {
+      "epoch": 0.6464603645612548,
+      "grad_norm": 1.901523232460022,
+      "learning_rate": 7.075973811878856e-05,
+      "loss": 2.0676,
+      "step": 27450
+    },
+    {
+      "epoch": 0.646695869247798,
+      "grad_norm": 2.3312385082244873,
+      "learning_rate": 7.071263718147991e-05,
+      "loss": 2.1138,
+      "step": 27460
+    },
+    {
+      "epoch": 0.6469313739343413,
+      "grad_norm": 2.5286500453948975,
+      "learning_rate": 7.066553624417126e-05,
+      "loss": 2.1407,
+      "step": 27470
+    },
+    {
+      "epoch": 0.6471668786208845,
+      "grad_norm": 1.9842777252197266,
+      "learning_rate": 7.062314540059347e-05,
+      "loss": 1.928,
+      "step": 27480
+    },
+    {
+      "epoch": 0.6474023833074278,
+      "grad_norm": 2.1205496788024902,
+      "learning_rate": 7.057604446328482e-05,
+      "loss": 2.0597,
+      "step": 27490
+    },
+    {
+      "epoch": 0.647637887993971,
+      "grad_norm": 2.428333044052124,
+      "learning_rate": 7.052894352597617e-05,
+      "loss": 2.0553,
+      "step": 27500
+    },
+    {
+      "epoch": 0.6478733926805144,
+      "grad_norm": 2.098172664642334,
+      "learning_rate": 7.048184258866752e-05,
+      "loss": 2.135,
+      "step": 27510
+    },
+    {
+      "epoch": 0.6481088973670576,
+      "grad_norm": 2.455638885498047,
+      "learning_rate": 7.043474165135887e-05,
+      "loss": 2.1021,
+      "step": 27520
+    },
+    {
+      "epoch": 0.6483444020536009,
+      "grad_norm": 1.9910706281661987,
+      "learning_rate": 7.038764071405022e-05,
+      "loss": 2.0287,
+      "step": 27530
+    },
+    {
+      "epoch": 0.6485799067401441,
+      "grad_norm": 1.8840402364730835,
+      "learning_rate": 7.034053977674157e-05,
+      "loss": 2.1737,
+      "step": 27540
+    },
+    {
+      "epoch": 0.6488154114266874,
+      "grad_norm": 2.0080809593200684,
+      "learning_rate": 7.029343883943292e-05,
+      "loss": 1.8489,
+      "step": 27550
+    },
+    {
+      "epoch": 0.6490509161132306,
+      "grad_norm": 2.091917037963867,
+      "learning_rate": 7.024633790212425e-05,
+      "loss": 2.0131,
+      "step": 27560
+    },
+    {
+      "epoch": 0.649286420799774,
+      "grad_norm": 1.8030489683151245,
+      "learning_rate": 7.01992369648156e-05,
+      "loss": 2.0919,
+      "step": 27570
+    },
+    {
+      "epoch": 0.6495219254863172,
+      "grad_norm": 1.896256923675537,
+      "learning_rate": 7.015213602750696e-05,
+      "loss": 2.0142,
+      "step": 27580
+    },
+    {
+      "epoch": 0.6497574301728605,
+      "grad_norm": 1.7844651937484741,
+      "learning_rate": 7.01050350901983e-05,
+      "loss": 2.2077,
+      "step": 27590
+    },
+    {
+      "epoch": 0.6499929348594037,
+      "grad_norm": 2.182560920715332,
+      "learning_rate": 7.005793415288964e-05,
+      "loss": 1.7973,
+      "step": 27600
+    },
+    {
+      "epoch": 0.650228439545947,
+      "grad_norm": 2.0249404907226562,
+      "learning_rate": 7.001083321558099e-05,
+      "loss": 2.1121,
+      "step": 27610
+    },
+    {
+      "epoch": 0.6504639442324902,
+      "grad_norm": 2.052584409713745,
+      "learning_rate": 6.996373227827234e-05,
+      "loss": 2.2913,
+      "step": 27620
+    },
+    {
+      "epoch": 0.6506994489190335,
+      "grad_norm": 1.8792239427566528,
+      "learning_rate": 6.991663134096369e-05,
+      "loss": 2.059,
+      "step": 27630
+    },
+    {
+      "epoch": 0.6509349536055767,
+      "grad_norm": 2.087036371231079,
+      "learning_rate": 6.986953040365504e-05,
+      "loss": 2.0064,
+      "step": 27640
+    },
+    {
+      "epoch": 0.6511704582921201,
+      "grad_norm": 2.4446024894714355,
+      "learning_rate": 6.982242946634638e-05,
+      "loss": 2.0678,
+      "step": 27650
+    },
+    {
+      "epoch": 0.6514059629786633,
+      "grad_norm": 2.256411552429199,
+      "learning_rate": 6.977532852903773e-05,
+      "loss": 1.9454,
+      "step": 27660
+    },
+    {
+      "epoch": 0.6516414676652066,
+      "grad_norm": 2.468778610229492,
+      "learning_rate": 6.972822759172908e-05,
+      "loss": 2.1624,
+      "step": 27670
+    },
+    {
+      "epoch": 0.6518769723517498,
+      "grad_norm": 2.614762544631958,
+      "learning_rate": 6.968112665442043e-05,
+      "loss": 2.1327,
+      "step": 27680
+    },
+    {
+      "epoch": 0.6521124770382931,
+      "grad_norm": 1.6540695428848267,
+      "learning_rate": 6.963402571711177e-05,
+      "loss": 1.7609,
+      "step": 27690
+    },
+    {
+      "epoch": 0.6523479817248363,
+      "grad_norm": 1.6981443166732788,
+      "learning_rate": 6.958692477980312e-05,
+      "loss": 1.9458,
+      "step": 27700
+    },
+    {
+      "epoch": 0.6525834864113795,
+      "grad_norm": 2.1197662353515625,
+      "learning_rate": 6.953982384249447e-05,
+      "loss": 2.1603,
+      "step": 27710
+    },
+    {
+      "epoch": 0.6528189910979229,
+      "grad_norm": 2.753307580947876,
+      "learning_rate": 6.949272290518582e-05,
+      "loss": 2.192,
+      "step": 27720
+    },
+    {
+      "epoch": 0.6530544957844661,
+      "grad_norm": 1.7629789113998413,
+      "learning_rate": 6.944562196787717e-05,
+      "loss": 1.9333,
+      "step": 27730
+    },
+    {
+      "epoch": 0.6532900004710094,
+      "grad_norm": 2.513230323791504,
+      "learning_rate": 6.93985210305685e-05,
+      "loss": 2.0639,
+      "step": 27740
+    },
+    {
+      "epoch": 0.6535255051575526,
+      "grad_norm": 1.9997378587722778,
+      "learning_rate": 6.935142009325986e-05,
+      "loss": 1.8298,
+      "step": 27750
+    },
+    {
+      "epoch": 0.6537610098440959,
+      "grad_norm": 2.183000087738037,
+      "learning_rate": 6.93043191559512e-05,
+      "loss": 1.8049,
+      "step": 27760
+    },
+    {
+      "epoch": 0.6539965145306391,
+      "grad_norm": 2.316758155822754,
+      "learning_rate": 6.925721821864256e-05,
+      "loss": 2.2343,
+      "step": 27770
+    },
+    {
+      "epoch": 0.6542320192171824,
+      "grad_norm": 2.8688292503356934,
+      "learning_rate": 6.921011728133389e-05,
+      "loss": 2.0771,
+      "step": 27780
+    },
+    {
+      "epoch": 0.6544675239037256,
+      "grad_norm": 2.846386671066284,
+      "learning_rate": 6.916301634402524e-05,
+      "loss": 1.9481,
+      "step": 27790
+    },
+    {
+      "epoch": 0.654703028590269,
+      "grad_norm": 2.241668462753296,
+      "learning_rate": 6.911591540671659e-05,
+      "loss": 2.0466,
+      "step": 27800
+    },
+    {
+      "epoch": 0.6549385332768122,
+      "grad_norm": 1.862688422203064,
+      "learning_rate": 6.906881446940796e-05,
+      "loss": 2.0278,
+      "step": 27810
+    },
+    {
+      "epoch": 0.6551740379633555,
+      "grad_norm": 2.669524908065796,
+      "learning_rate": 6.902171353209929e-05,
+      "loss": 1.9347,
+      "step": 27820
+    },
+    {
+      "epoch": 0.6554095426498987,
+      "grad_norm": 1.9434707164764404,
+      "learning_rate": 6.897461259479064e-05,
+      "loss": 1.9777,
+      "step": 27830
+    },
+    {
+      "epoch": 0.655645047336442,
+      "grad_norm": 1.9469919204711914,
+      "learning_rate": 6.8927511657482e-05,
+      "loss": 2.1351,
+      "step": 27840
+    },
+    {
+      "epoch": 0.6558805520229852,
+      "grad_norm": 2.0823862552642822,
+      "learning_rate": 6.888041072017334e-05,
+      "loss": 2.1103,
+      "step": 27850
+    },
+    {
+      "epoch": 0.6561160567095286,
+      "grad_norm": 2.011646270751953,
+      "learning_rate": 6.883330978286468e-05,
+      "loss": 2.0807,
+      "step": 27860
+    },
+    {
+      "epoch": 0.6563515613960718,
+      "grad_norm": 2.085239887237549,
+      "learning_rate": 6.878620884555603e-05,
+      "loss": 2.0131,
+      "step": 27870
+    },
+    {
+      "epoch": 0.6565870660826151,
+      "grad_norm": 2.445866346359253,
+      "learning_rate": 6.873910790824738e-05,
+      "loss": 1.9386,
+      "step": 27880
+    },
+    {
+      "epoch": 0.6568225707691583,
+      "grad_norm": 2.3337929248809814,
+      "learning_rate": 6.869200697093873e-05,
+      "loss": 1.9866,
+      "step": 27890
+    },
+    {
+      "epoch": 0.6570580754557016,
+      "grad_norm": 2.558706045150757,
+      "learning_rate": 6.864490603363007e-05,
+      "loss": 1.9361,
+      "step": 27900
+    },
+    {
+      "epoch": 0.6572935801422448,
+      "grad_norm": 1.7763642072677612,
+      "learning_rate": 6.859780509632142e-05,
+      "loss": 2.147,
+      "step": 27910
+    },
+    {
+      "epoch": 0.6575290848287881,
+      "grad_norm": 2.353127956390381,
+      "learning_rate": 6.855070415901277e-05,
+      "loss": 1.972,
+      "step": 27920
+    },
+    {
+      "epoch": 0.6577645895153313,
+      "grad_norm": 2.1729345321655273,
+      "learning_rate": 6.850360322170412e-05,
+      "loss": 1.9824,
+      "step": 27930
+    },
+    {
+      "epoch": 0.6580000942018747,
+      "grad_norm": 2.3141558170318604,
+      "learning_rate": 6.845650228439547e-05,
+      "loss": 1.8974,
+      "step": 27940
+    },
+    {
+      "epoch": 0.6582355988884179,
+      "grad_norm": 2.9359517097473145,
+      "learning_rate": 6.84094013470868e-05,
+      "loss": 2.0506,
+      "step": 27950
+    },
+    {
+      "epoch": 0.6584711035749612,
+      "grad_norm": 2.285006046295166,
+      "learning_rate": 6.836230040977816e-05,
+      "loss": 2.1571,
+      "step": 27960
+    },
+    {
+      "epoch": 0.6587066082615044,
+      "grad_norm": 2.683715581893921,
+      "learning_rate": 6.83151994724695e-05,
+      "loss": 2.3031,
+      "step": 27970
+    },
+    {
+      "epoch": 0.6589421129480477,
+      "grad_norm": 2.042071580886841,
+      "learning_rate": 6.826809853516086e-05,
+      "loss": 1.9662,
+      "step": 27980
+    },
+    {
+      "epoch": 0.6591776176345909,
+      "grad_norm": 2.244554042816162,
+      "learning_rate": 6.822099759785219e-05,
+      "loss": 2.1763,
+      "step": 27990
+    },
+    {
+      "epoch": 0.6594131223211341,
+      "grad_norm": 2.151461124420166,
+      "learning_rate": 6.817389666054354e-05,
+      "loss": 1.8774,
+      "step": 28000
+    },
+    {
+      "epoch": 0.6596486270076775,
+      "grad_norm": 2.07513427734375,
+      "learning_rate": 6.81267957232349e-05,
+      "loss": 1.9599,
+      "step": 28010
+    },
+    {
+      "epoch": 0.6598841316942207,
+      "grad_norm": 1.9893085956573486,
+      "learning_rate": 6.807969478592624e-05,
+      "loss": 2.1541,
+      "step": 28020
+    },
+    {
+      "epoch": 0.660119636380764,
+      "grad_norm": 3.486293315887451,
+      "learning_rate": 6.80325938486176e-05,
+      "loss": 2.1084,
+      "step": 28030
+    },
+    {
+      "epoch": 0.6603551410673072,
+      "grad_norm": 2.266904830932617,
+      "learning_rate": 6.798549291130893e-05,
+      "loss": 1.9884,
+      "step": 28040
+    },
+    {
+      "epoch": 0.6605906457538505,
+      "grad_norm": 2.0467066764831543,
+      "learning_rate": 6.793839197400028e-05,
+      "loss": 2.172,
+      "step": 28050
+    },
+    {
+      "epoch": 0.6608261504403937,
+      "grad_norm": 2.1317572593688965,
+      "learning_rate": 6.789129103669163e-05,
+      "loss": 2.1142,
+      "step": 28060
+    },
+    {
+      "epoch": 0.661061655126937,
+      "grad_norm": 2.3432118892669678,
+      "learning_rate": 6.784419009938298e-05,
+      "loss": 1.9512,
+      "step": 28070
+    },
+    {
+      "epoch": 0.6612971598134803,
+      "grad_norm": 2.4152426719665527,
+      "learning_rate": 6.779708916207432e-05,
+      "loss": 2.0989,
+      "step": 28080
+    },
+    {
+      "epoch": 0.6615326645000236,
+      "grad_norm": 1.8867883682250977,
+      "learning_rate": 6.774998822476568e-05,
+      "loss": 1.9836,
+      "step": 28090
+    },
+    {
+      "epoch": 0.6617681691865668,
+      "grad_norm": 2.182126998901367,
+      "learning_rate": 6.770288728745703e-05,
+      "loss": 1.8761,
+      "step": 28100
+    },
+    {
+      "epoch": 0.6620036738731101,
+      "grad_norm": 2.5316390991210938,
+      "learning_rate": 6.765578635014838e-05,
+      "loss": 1.9713,
+      "step": 28110
+    },
+    {
+      "epoch": 0.6622391785596533,
+      "grad_norm": 2.1727163791656494,
+      "learning_rate": 6.760868541283972e-05,
+      "loss": 1.8554,
+      "step": 28120
+    },
+    {
+      "epoch": 0.6624746832461966,
+      "grad_norm": 3.6611173152923584,
+      "learning_rate": 6.756158447553107e-05,
+      "loss": 1.9843,
+      "step": 28130
+    },
+    {
+      "epoch": 0.6627101879327398,
+      "grad_norm": 2.4131314754486084,
+      "learning_rate": 6.751448353822242e-05,
+      "loss": 2.0242,
+      "step": 28140
+    },
+    {
+      "epoch": 0.6629456926192832,
+      "grad_norm": 1.8160946369171143,
+      "learning_rate": 6.746738260091377e-05,
+      "loss": 2.0622,
+      "step": 28150
+    },
+    {
+      "epoch": 0.6631811973058264,
+      "grad_norm": 3.188413381576538,
+      "learning_rate": 6.74202816636051e-05,
+      "loss": 2.2114,
+      "step": 28160
+    },
+    {
+      "epoch": 0.6634167019923697,
+      "grad_norm": 2.4607508182525635,
+      "learning_rate": 6.737318072629646e-05,
+      "loss": 2.0314,
+      "step": 28170
+    },
+    {
+      "epoch": 0.6636522066789129,
+      "grad_norm": 2.2782578468322754,
+      "learning_rate": 6.732607978898781e-05,
+      "loss": 1.9546,
+      "step": 28180
+    },
+    {
+      "epoch": 0.6638877113654562,
+      "grad_norm": 1.997504472732544,
+      "learning_rate": 6.727897885167916e-05,
+      "loss": 2.0726,
+      "step": 28190
+    },
+    {
+      "epoch": 0.6641232160519994,
+      "grad_norm": 2.3978195190429688,
+      "learning_rate": 6.72318779143705e-05,
+      "loss": 2.1611,
+      "step": 28200
+    },
+    {
+      "epoch": 0.6643587207385427,
+      "grad_norm": 2.00547456741333,
+      "learning_rate": 6.718477697706184e-05,
+      "loss": 2.0716,
+      "step": 28210
+    },
+    {
+      "epoch": 0.664594225425086,
+      "grad_norm": 1.96976637840271,
+      "learning_rate": 6.71376760397532e-05,
+      "loss": 1.9797,
+      "step": 28220
+    },
+    {
+      "epoch": 0.6648297301116293,
+      "grad_norm": 2.197422981262207,
+      "learning_rate": 6.709057510244454e-05,
+      "loss": 1.9887,
+      "step": 28230
+    },
+    {
+      "epoch": 0.6650652347981725,
+      "grad_norm": 1.326515555381775,
+      "learning_rate": 6.70434741651359e-05,
+      "loss": 1.9457,
+      "step": 28240
+    },
+    {
+      "epoch": 0.6653007394847158,
+      "grad_norm": 1.9585696458816528,
+      "learning_rate": 6.699637322782723e-05,
+      "loss": 2.1483,
+      "step": 28250
+    },
+    {
+      "epoch": 0.665536244171259,
+      "grad_norm": 2.264357328414917,
+      "learning_rate": 6.694927229051858e-05,
+      "loss": 1.9455,
+      "step": 28260
+    },
+    {
+      "epoch": 0.6657717488578023,
+      "grad_norm": 2.1170806884765625,
+      "learning_rate": 6.690217135320993e-05,
+      "loss": 2.1229,
+      "step": 28270
+    },
+    {
+      "epoch": 0.6660072535443455,
+      "grad_norm": 2.3567802906036377,
+      "learning_rate": 6.685507041590128e-05,
+      "loss": 2.2026,
+      "step": 28280
+    },
+    {
+      "epoch": 0.6662427582308887,
+      "grad_norm": 2.09476637840271,
+      "learning_rate": 6.680796947859262e-05,
+      "loss": 1.9125,
+      "step": 28290
+    },
+    {
+      "epoch": 0.6664782629174321,
+      "grad_norm": 1.959568738937378,
+      "learning_rate": 6.676086854128397e-05,
+      "loss": 2.085,
+      "step": 28300
+    },
+    {
+      "epoch": 0.6667137676039753,
+      "grad_norm": 3.108013153076172,
+      "learning_rate": 6.671376760397532e-05,
+      "loss": 1.9585,
+      "step": 28310
+    },
+    {
+      "epoch": 0.6669492722905186,
+      "grad_norm": 2.3948521614074707,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 1.8998,
+      "step": 28320
+    },
+    {
+      "epoch": 0.6671847769770618,
+      "grad_norm": 1.9991596937179565,
+      "learning_rate": 6.6619565729358e-05,
+      "loss": 1.9824,
+      "step": 28330
+    },
+    {
+      "epoch": 0.6674202816636051,
+      "grad_norm": 1.759979486465454,
+      "learning_rate": 6.657246479204936e-05,
+      "loss": 1.9744,
+      "step": 28340
+    },
+    {
+      "epoch": 0.6676557863501483,
+      "grad_norm": 2.247377634048462,
+      "learning_rate": 6.652536385474071e-05,
+      "loss": 2.1226,
+      "step": 28350
+    },
+    {
+      "epoch": 0.6678912910366916,
+      "grad_norm": 3.3564887046813965,
+      "learning_rate": 6.647826291743206e-05,
+      "loss": 2.0988,
+      "step": 28360
+    },
+    {
+      "epoch": 0.6681267957232349,
+      "grad_norm": 2.1177890300750732,
+      "learning_rate": 6.643116198012341e-05,
+      "loss": 1.6984,
+      "step": 28370
+    },
+    {
+      "epoch": 0.6683623004097782,
+      "grad_norm": 2.21870756149292,
+      "learning_rate": 6.638406104281476e-05,
+      "loss": 2.1458,
+      "step": 28380
+    },
+    {
+      "epoch": 0.6685978050963214,
+      "grad_norm": 1.8025680780410767,
+      "learning_rate": 6.633696010550611e-05,
+      "loss": 1.9258,
+      "step": 28390
+    },
+    {
+      "epoch": 0.6688333097828647,
+      "grad_norm": 1.9146382808685303,
+      "learning_rate": 6.628985916819746e-05,
+      "loss": 1.879,
+      "step": 28400
+    },
+    {
+      "epoch": 0.6690688144694079,
+      "grad_norm": 2.1155905723571777,
+      "learning_rate": 6.62427582308888e-05,
+      "loss": 2.2079,
+      "step": 28410
+    },
+    {
+      "epoch": 0.6693043191559512,
+      "grad_norm": 2.081460475921631,
+      "learning_rate": 6.619565729358015e-05,
+      "loss": 2.0226,
+      "step": 28420
+    },
+    {
+      "epoch": 0.6695398238424944,
+      "grad_norm": 2.065375328063965,
+      "learning_rate": 6.61485563562715e-05,
+      "loss": 2.0476,
+      "step": 28430
+    },
+    {
+      "epoch": 0.6697753285290378,
+      "grad_norm": 1.6890816688537598,
+      "learning_rate": 6.610145541896285e-05,
+      "loss": 1.9391,
+      "step": 28440
+    },
+    {
+      "epoch": 0.670010833215581,
+      "grad_norm": 2.4403939247131348,
+      "learning_rate": 6.60543544816542e-05,
+      "loss": 1.8528,
+      "step": 28450
+    },
+    {
+      "epoch": 0.6702463379021243,
+      "grad_norm": 2.6802732944488525,
+      "learning_rate": 6.600725354434553e-05,
+      "loss": 2.1087,
+      "step": 28460
+    },
+    {
+      "epoch": 0.6704818425886675,
+      "grad_norm": 1.8135290145874023,
+      "learning_rate": 6.596015260703688e-05,
+      "loss": 2.1068,
+      "step": 28470
+    },
+    {
+      "epoch": 0.6707173472752108,
+      "grad_norm": 2.0007054805755615,
+      "learning_rate": 6.591305166972823e-05,
+      "loss": 2.113,
+      "step": 28480
+    },
+    {
+      "epoch": 0.670952851961754,
+      "grad_norm": 2.349604368209839,
+      "learning_rate": 6.586595073241958e-05,
+      "loss": 1.7037,
+      "step": 28490
+    },
+    {
+      "epoch": 0.6711883566482973,
+      "grad_norm": 2.2065269947052,
+      "learning_rate": 6.581884979511092e-05,
+      "loss": 2.3166,
+      "step": 28500
+    },
+    {
+      "epoch": 0.6714238613348406,
+      "grad_norm": 2.0659828186035156,
+      "learning_rate": 6.577174885780227e-05,
+      "loss": 2.0392,
+      "step": 28510
+    },
+    {
+      "epoch": 0.6716593660213839,
+      "grad_norm": 2.7772090435028076,
+      "learning_rate": 6.572464792049362e-05,
+      "loss": 1.9277,
+      "step": 28520
+    },
+    {
+      "epoch": 0.6718948707079271,
+      "grad_norm": 2.1206893920898438,
+      "learning_rate": 6.567754698318497e-05,
+      "loss": 1.8924,
+      "step": 28530
+    },
+    {
+      "epoch": 0.6721303753944704,
+      "grad_norm": 2.0614333152770996,
+      "learning_rate": 6.563044604587632e-05,
+      "loss": 2.1831,
+      "step": 28540
+    },
+    {
+      "epoch": 0.6723658800810136,
+      "grad_norm": 1.9979201555252075,
+      "learning_rate": 6.558334510856766e-05,
+      "loss": 1.9522,
+      "step": 28550
+    },
+    {
+      "epoch": 0.6726013847675568,
+      "grad_norm": 1.968994379043579,
+      "learning_rate": 6.553624417125901e-05,
+      "loss": 1.9722,
+      "step": 28560
+    },
+    {
+      "epoch": 0.6728368894541001,
+      "grad_norm": 1.783308982849121,
+      "learning_rate": 6.548914323395036e-05,
+      "loss": 2.1729,
+      "step": 28570
+    },
+    {
+      "epoch": 0.6730723941406433,
+      "grad_norm": 2.494913101196289,
+      "learning_rate": 6.544204229664171e-05,
+      "loss": 2.1268,
+      "step": 28580
+    },
+    {
+      "epoch": 0.6733078988271867,
+      "grad_norm": 2.0739006996154785,
+      "learning_rate": 6.539494135933305e-05,
+      "loss": 2.0764,
+      "step": 28590
+    },
+    {
+      "epoch": 0.6735434035137299,
+      "grad_norm": 3.6099002361297607,
+      "learning_rate": 6.53478404220244e-05,
+      "loss": 2.0012,
+      "step": 28600
+    },
+    {
+      "epoch": 0.6737789082002732,
+      "grad_norm": 2.3188111782073975,
+      "learning_rate": 6.530073948471575e-05,
+      "loss": 2.0252,
+      "step": 28610
+    },
+    {
+      "epoch": 0.6740144128868164,
+      "grad_norm": 1.7235386371612549,
+      "learning_rate": 6.525834864113795e-05,
+      "loss": 1.9303,
+      "step": 28620
+    },
+    {
+      "epoch": 0.6742499175733597,
+      "grad_norm": 1.7218267917633057,
+      "learning_rate": 6.521124770382931e-05,
+      "loss": 1.9484,
+      "step": 28630
+    },
+    {
+      "epoch": 0.6744854222599029,
+      "grad_norm": 2.311770439147949,
+      "learning_rate": 6.516414676652066e-05,
+      "loss": 2.0987,
+      "step": 28640
+    },
+    {
+      "epoch": 0.6747209269464463,
+      "grad_norm": 2.6366100311279297,
+      "learning_rate": 6.511704582921201e-05,
+      "loss": 1.9514,
+      "step": 28650
+    },
+    {
+      "epoch": 0.6749564316329895,
+      "grad_norm": 2.66902494430542,
+      "learning_rate": 6.506994489190335e-05,
+      "loss": 1.9368,
+      "step": 28660
+    },
+    {
+      "epoch": 0.6751919363195328,
+      "grad_norm": 2.354694366455078,
+      "learning_rate": 6.50228439545947e-05,
+      "loss": 1.9255,
+      "step": 28670
+    },
+    {
+      "epoch": 0.675427441006076,
+      "grad_norm": 2.0443005561828613,
+      "learning_rate": 6.497574301728605e-05,
+      "loss": 2.0929,
+      "step": 28680
+    },
+    {
+      "epoch": 0.6756629456926193,
+      "grad_norm": 1.8448538780212402,
+      "learning_rate": 6.49286420799774e-05,
+      "loss": 2.0847,
+      "step": 28690
+    },
+    {
+      "epoch": 0.6758984503791625,
+      "grad_norm": 2.754966974258423,
+      "learning_rate": 6.488154114266874e-05,
+      "loss": 2.1979,
+      "step": 28700
+    },
+    {
+      "epoch": 0.6761339550657058,
+      "grad_norm": 2.3212506771087646,
+      "learning_rate": 6.483444020536009e-05,
+      "loss": 2.062,
+      "step": 28710
+    },
+    {
+      "epoch": 0.676369459752249,
+      "grad_norm": 2.090217351913452,
+      "learning_rate": 6.478733926805144e-05,
+      "loss": 1.9911,
+      "step": 28720
+    },
+    {
+      "epoch": 0.6766049644387924,
+      "grad_norm": 1.9577314853668213,
+      "learning_rate": 6.474023833074279e-05,
+      "loss": 1.9376,
+      "step": 28730
+    },
+    {
+      "epoch": 0.6768404691253356,
+      "grad_norm": 1.8566160202026367,
+      "learning_rate": 6.469313739343414e-05,
+      "loss": 2.0811,
+      "step": 28740
+    },
+    {
+      "epoch": 0.6770759738118789,
+      "grad_norm": 2.3416032791137695,
+      "learning_rate": 6.464603645612548e-05,
+      "loss": 2.1295,
+      "step": 28750
+    },
+    {
+      "epoch": 0.6773114784984221,
+      "grad_norm": 1.6582307815551758,
+      "learning_rate": 6.459893551881683e-05,
+      "loss": 2.0106,
+      "step": 28760
+    },
+    {
+      "epoch": 0.6775469831849654,
+      "grad_norm": 2.144157886505127,
+      "learning_rate": 6.455183458150818e-05,
+      "loss": 1.9546,
+      "step": 28770
+    },
+    {
+      "epoch": 0.6777824878715086,
+      "grad_norm": 2.9796035289764404,
+      "learning_rate": 6.450473364419953e-05,
+      "loss": 1.9336,
+      "step": 28780
+    },
+    {
+      "epoch": 0.678017992558052,
+      "grad_norm": 2.0936977863311768,
+      "learning_rate": 6.445763270689086e-05,
+      "loss": 2.077,
+      "step": 28790
+    },
+    {
+      "epoch": 0.6782534972445952,
+      "grad_norm": 1.9568135738372803,
+      "learning_rate": 6.441053176958221e-05,
+      "loss": 2.1443,
+      "step": 28800
+    },
+    {
+      "epoch": 0.6784890019311385,
+      "grad_norm": 2.607478618621826,
+      "learning_rate": 6.436343083227356e-05,
+      "loss": 2.139,
+      "step": 28810
+    },
+    {
+      "epoch": 0.6787245066176817,
+      "grad_norm": 2.0148494243621826,
+      "learning_rate": 6.431632989496491e-05,
+      "loss": 1.7023,
+      "step": 28820
+    },
+    {
+      "epoch": 0.678960011304225,
+      "grad_norm": 2.429159164428711,
+      "learning_rate": 6.426922895765626e-05,
+      "loss": 1.9094,
+      "step": 28830
+    },
+    {
+      "epoch": 0.6791955159907682,
+      "grad_norm": 2.1411314010620117,
+      "learning_rate": 6.42221280203476e-05,
+      "loss": 2.1745,
+      "step": 28840
+    },
+    {
+      "epoch": 0.6794310206773114,
+      "grad_norm": 2.190335988998413,
+      "learning_rate": 6.417502708303895e-05,
+      "loss": 2.139,
+      "step": 28850
+    },
+    {
+      "epoch": 0.6796665253638547,
+      "grad_norm": 2.283425807952881,
+      "learning_rate": 6.41279261457303e-05,
+      "loss": 1.9698,
+      "step": 28860
+    },
+    {
+      "epoch": 0.679902030050398,
+      "grad_norm": 2.0648112297058105,
+      "learning_rate": 6.408082520842165e-05,
+      "loss": 2.0706,
+      "step": 28870
+    },
+    {
+      "epoch": 0.6801375347369413,
+      "grad_norm": 1.8722620010375977,
+      "learning_rate": 6.403372427111299e-05,
+      "loss": 2.0114,
+      "step": 28880
+    },
+    {
+      "epoch": 0.6803730394234845,
+      "grad_norm": 2.4573814868927,
+      "learning_rate": 6.398662333380434e-05,
+      "loss": 1.9978,
+      "step": 28890
+    },
+    {
+      "epoch": 0.6806085441100278,
+      "grad_norm": 2.1509108543395996,
+      "learning_rate": 6.393952239649569e-05,
+      "loss": 2.0784,
+      "step": 28900
+    },
+    {
+      "epoch": 0.680844048796571,
+      "grad_norm": 2.1195876598358154,
+      "learning_rate": 6.389242145918705e-05,
+      "loss": 2.1025,
+      "step": 28910
+    },
+    {
+      "epoch": 0.6810795534831143,
+      "grad_norm": 2.044261932373047,
+      "learning_rate": 6.384532052187839e-05,
+      "loss": 1.786,
+      "step": 28920
+    },
+    {
+      "epoch": 0.6813150581696575,
+      "grad_norm": 3.030416250228882,
+      "learning_rate": 6.379821958456974e-05,
+      "loss": 2.1121,
+      "step": 28930
+    },
+    {
+      "epoch": 0.6815505628562009,
+      "grad_norm": 2.920388698577881,
+      "learning_rate": 6.375111864726109e-05,
+      "loss": 2.0764,
+      "step": 28940
+    },
+    {
+      "epoch": 0.6817860675427441,
+      "grad_norm": 1.9454491138458252,
+      "learning_rate": 6.370401770995244e-05,
+      "loss": 2.0691,
+      "step": 28950
+    },
+    {
+      "epoch": 0.6820215722292874,
+      "grad_norm": 1.834728479385376,
+      "learning_rate": 6.365691677264378e-05,
+      "loss": 1.8553,
+      "step": 28960
+    },
+    {
+      "epoch": 0.6822570769158306,
+      "grad_norm": 2.1100876331329346,
+      "learning_rate": 6.360981583533513e-05,
+      "loss": 1.8956,
+      "step": 28970
+    },
+    {
+      "epoch": 0.6824925816023739,
+      "grad_norm": 1.9620522260665894,
+      "learning_rate": 6.356271489802648e-05,
+      "loss": 1.7718,
+      "step": 28980
+    },
+    {
+      "epoch": 0.6827280862889171,
+      "grad_norm": 2.126603126525879,
+      "learning_rate": 6.351561396071783e-05,
+      "loss": 1.9046,
+      "step": 28990
+    },
+    {
+      "epoch": 0.6829635909754604,
+      "grad_norm": 2.21046781539917,
+      "learning_rate": 6.346851302340916e-05,
+      "loss": 2.054,
+      "step": 29000
+    },
+    {
+      "epoch": 0.6831990956620037,
+      "grad_norm": 2.2389838695526123,
+      "learning_rate": 6.342141208610051e-05,
+      "loss": 2.0818,
+      "step": 29010
+    },
+    {
+      "epoch": 0.683434600348547,
+      "grad_norm": 2.375826358795166,
+      "learning_rate": 6.337431114879186e-05,
+      "loss": 1.9759,
+      "step": 29020
+    },
+    {
+      "epoch": 0.6836701050350902,
+      "grad_norm": 1.9584667682647705,
+      "learning_rate": 6.332721021148322e-05,
+      "loss": 2.0202,
+      "step": 29030
+    },
+    {
+      "epoch": 0.6839056097216335,
+      "grad_norm": 2.5491538047790527,
+      "learning_rate": 6.328010927417457e-05,
+      "loss": 2.0253,
+      "step": 29040
+    },
+    {
+      "epoch": 0.6841411144081767,
+      "grad_norm": 1.8970832824707031,
+      "learning_rate": 6.32330083368659e-05,
+      "loss": 1.842,
+      "step": 29050
+    },
+    {
+      "epoch": 0.68437661909472,
+      "grad_norm": 2.02691912651062,
+      "learning_rate": 6.318590739955725e-05,
+      "loss": 1.788,
+      "step": 29060
+    },
+    {
+      "epoch": 0.6846121237812632,
+      "grad_norm": 1.954071283340454,
+      "learning_rate": 6.31388064622486e-05,
+      "loss": 1.9792,
+      "step": 29070
+    },
+    {
+      "epoch": 0.6848476284678066,
+      "grad_norm": 2.6439526081085205,
+      "learning_rate": 6.309170552493995e-05,
+      "loss": 1.8697,
+      "step": 29080
+    },
+    {
+      "epoch": 0.6850831331543498,
+      "grad_norm": 2.4627387523651123,
+      "learning_rate": 6.304460458763129e-05,
+      "loss": 2.2861,
+      "step": 29090
+    },
+    {
+      "epoch": 0.6853186378408931,
+      "grad_norm": 2.732792615890503,
+      "learning_rate": 6.299750365032264e-05,
+      "loss": 1.9502,
+      "step": 29100
+    },
+    {
+      "epoch": 0.6855541425274363,
+      "grad_norm": 2.2963614463806152,
+      "learning_rate": 6.295040271301399e-05,
+      "loss": 2.0442,
+      "step": 29110
+    },
+    {
+      "epoch": 0.6857896472139795,
+      "grad_norm": 2.23295259475708,
+      "learning_rate": 6.290330177570534e-05,
+      "loss": 2.2432,
+      "step": 29120
+    },
+    {
+      "epoch": 0.6860251519005228,
+      "grad_norm": 2.274610996246338,
+      "learning_rate": 6.285620083839669e-05,
+      "loss": 2.1089,
+      "step": 29130
+    },
+    {
+      "epoch": 0.686260656587066,
+      "grad_norm": 2.2117106914520264,
+      "learning_rate": 6.280909990108803e-05,
+      "loss": 2.1116,
+      "step": 29140
+    },
+    {
+      "epoch": 0.6864961612736094,
+      "grad_norm": 1.9509594440460205,
+      "learning_rate": 6.276199896377938e-05,
+      "loss": 2.033,
+      "step": 29150
+    },
+    {
+      "epoch": 0.6867316659601526,
+      "grad_norm": 2.090939521789551,
+      "learning_rate": 6.271489802647073e-05,
+      "loss": 2.1308,
+      "step": 29160
+    },
+    {
+      "epoch": 0.6869671706466959,
+      "grad_norm": 1.9330859184265137,
+      "learning_rate": 6.266779708916208e-05,
+      "loss": 2.0138,
+      "step": 29170
+    },
+    {
+      "epoch": 0.6872026753332391,
+      "grad_norm": 2.7607204914093018,
+      "learning_rate": 6.262069615185341e-05,
+      "loss": 1.8859,
+      "step": 29180
+    },
+    {
+      "epoch": 0.6874381800197824,
+      "grad_norm": 1.8978123664855957,
+      "learning_rate": 6.257359521454478e-05,
+      "loss": 1.9501,
+      "step": 29190
+    },
+    {
+      "epoch": 0.6876736847063256,
+      "grad_norm": 1.9100096225738525,
+      "learning_rate": 6.252649427723613e-05,
+      "loss": 2.1775,
+      "step": 29200
+    },
+    {
+      "epoch": 0.6879091893928689,
+      "grad_norm": 2.1695330142974854,
+      "learning_rate": 6.247939333992748e-05,
+      "loss": 1.8777,
+      "step": 29210
+    },
+    {
+      "epoch": 0.6881446940794121,
+      "grad_norm": 2.1711602210998535,
+      "learning_rate": 6.243229240261882e-05,
+      "loss": 1.9542,
+      "step": 29220
+    },
+    {
+      "epoch": 0.6883801987659555,
+      "grad_norm": 1.9122905731201172,
+      "learning_rate": 6.238519146531017e-05,
+      "loss": 2.0965,
+      "step": 29230
+    },
+    {
+      "epoch": 0.6886157034524987,
+      "grad_norm": 2.8225343227386475,
+      "learning_rate": 6.233809052800152e-05,
+      "loss": 1.8979,
+      "step": 29240
+    },
+    {
+      "epoch": 0.688851208139042,
+      "grad_norm": 2.511843204498291,
+      "learning_rate": 6.229098959069287e-05,
+      "loss": 1.8626,
+      "step": 29250
+    },
+    {
+      "epoch": 0.6890867128255852,
+      "grad_norm": 2.351067066192627,
+      "learning_rate": 6.22438886533842e-05,
+      "loss": 1.9113,
+      "step": 29260
+    },
+    {
+      "epoch": 0.6893222175121285,
+      "grad_norm": 2.117947578430176,
+      "learning_rate": 6.219678771607555e-05,
+      "loss": 2.2033,
+      "step": 29270
+    },
+    {
+      "epoch": 0.6895577221986717,
+      "grad_norm": 2.907036542892456,
+      "learning_rate": 6.21496867787669e-05,
+      "loss": 1.9599,
+      "step": 29280
+    },
+    {
+      "epoch": 0.689793226885215,
+      "grad_norm": 2.5090277194976807,
+      "learning_rate": 6.210258584145825e-05,
+      "loss": 2.0079,
+      "step": 29290
+    },
+    {
+      "epoch": 0.6900287315717583,
+      "grad_norm": 2.303168296813965,
+      "learning_rate": 6.205548490414959e-05,
+      "loss": 2.1558,
+      "step": 29300
+    },
+    {
+      "epoch": 0.6902642362583016,
+      "grad_norm": 1.9022759199142456,
+      "learning_rate": 6.200838396684094e-05,
+      "loss": 1.9063,
+      "step": 29310
+    },
+    {
+      "epoch": 0.6904997409448448,
+      "grad_norm": 2.1879916191101074,
+      "learning_rate": 6.196128302953229e-05,
+      "loss": 1.8982,
+      "step": 29320
+    },
+    {
+      "epoch": 0.6907352456313881,
+      "grad_norm": 2.6008474826812744,
+      "learning_rate": 6.191418209222364e-05,
+      "loss": 2.2362,
+      "step": 29330
+    },
+    {
+      "epoch": 0.6909707503179313,
+      "grad_norm": 2.61698317527771,
+      "learning_rate": 6.186708115491499e-05,
+      "loss": 2.0725,
+      "step": 29340
+    },
+    {
+      "epoch": 0.6912062550044746,
+      "grad_norm": 2.21352219581604,
+      "learning_rate": 6.181998021760633e-05,
+      "loss": 2.0445,
+      "step": 29350
+    },
+    {
+      "epoch": 0.6914417596910178,
+      "grad_norm": 2.037073850631714,
+      "learning_rate": 6.177287928029768e-05,
+      "loss": 1.8788,
+      "step": 29360
+    },
+    {
+      "epoch": 0.6916772643775612,
+      "grad_norm": 2.1956939697265625,
+      "learning_rate": 6.172577834298903e-05,
+      "loss": 2.0399,
+      "step": 29370
+    },
+    {
+      "epoch": 0.6919127690641044,
+      "grad_norm": 2.5270023345947266,
+      "learning_rate": 6.167867740568038e-05,
+      "loss": 2.2344,
+      "step": 29380
+    },
+    {
+      "epoch": 0.6921482737506477,
+      "grad_norm": 2.476280689239502,
+      "learning_rate": 6.163157646837172e-05,
+      "loss": 1.9711,
+      "step": 29390
+    },
+    {
+      "epoch": 0.6923837784371909,
+      "grad_norm": 1.4749599695205688,
+      "learning_rate": 6.158447553106307e-05,
+      "loss": 2.0054,
+      "step": 29400
+    },
+    {
+      "epoch": 0.6926192831237341,
+      "grad_norm": 2.379751443862915,
+      "learning_rate": 6.153737459375442e-05,
+      "loss": 1.9132,
+      "step": 29410
+    },
+    {
+      "epoch": 0.6928547878102774,
+      "grad_norm": 2.010408401489258,
+      "learning_rate": 6.149027365644577e-05,
+      "loss": 2.0497,
+      "step": 29420
+    },
+    {
+      "epoch": 0.6930902924968206,
+      "grad_norm": 1.9814579486846924,
+      "learning_rate": 6.144317271913712e-05,
+      "loss": 2.0795,
+      "step": 29430
+    },
+    {
+      "epoch": 0.693325797183364,
+      "grad_norm": 2.5224850177764893,
+      "learning_rate": 6.139607178182845e-05,
+      "loss": 1.8552,
+      "step": 29440
+    },
+    {
+      "epoch": 0.6935613018699072,
+      "grad_norm": 2.3376238346099854,
+      "learning_rate": 6.13489708445198e-05,
+      "loss": 1.7499,
+      "step": 29450
+    },
+    {
+      "epoch": 0.6937968065564505,
+      "grad_norm": 2.1768531799316406,
+      "learning_rate": 6.130186990721115e-05,
+      "loss": 2.1749,
+      "step": 29460
+    },
+    {
+      "epoch": 0.6940323112429937,
+      "grad_norm": 2.1843111515045166,
+      "learning_rate": 6.12547689699025e-05,
+      "loss": 2.0094,
+      "step": 29470
+    },
+    {
+      "epoch": 0.694267815929537,
+      "grad_norm": 2.0782411098480225,
+      "learning_rate": 6.120766803259385e-05,
+      "loss": 2.1589,
+      "step": 29480
+    },
+    {
+      "epoch": 0.6945033206160802,
+      "grad_norm": 2.4557056427001953,
+      "learning_rate": 6.11605670952852e-05,
+      "loss": 2.164,
+      "step": 29490
+    },
+    {
+      "epoch": 0.6947388253026235,
+      "grad_norm": 2.396660804748535,
+      "learning_rate": 6.111346615797655e-05,
+      "loss": 2.012,
+      "step": 29500
+    },
+    {
+      "epoch": 0.6949743299891668,
+      "grad_norm": 2.3271138668060303,
+      "learning_rate": 6.10663652206679e-05,
+      "loss": 1.9762,
+      "step": 29510
+    },
+    {
+      "epoch": 0.6952098346757101,
+      "grad_norm": 2.339451789855957,
+      "learning_rate": 6.101926428335925e-05,
+      "loss": 1.7365,
+      "step": 29520
+    },
+    {
+      "epoch": 0.6954453393622533,
+      "grad_norm": 1.9744147062301636,
+      "learning_rate": 6.097216334605059e-05,
+      "loss": 2.0507,
+      "step": 29530
+    },
+    {
+      "epoch": 0.6956808440487966,
+      "grad_norm": 3.988935708999634,
+      "learning_rate": 6.092506240874194e-05,
+      "loss": 1.9501,
+      "step": 29540
+    },
+    {
+      "epoch": 0.6959163487353398,
+      "grad_norm": 2.2221195697784424,
+      "learning_rate": 6.0877961471433286e-05,
+      "loss": 2.123,
+      "step": 29550
+    },
+    {
+      "epoch": 0.6961518534218831,
+      "grad_norm": 2.0011703968048096,
+      "learning_rate": 6.0830860534124636e-05,
+      "loss": 1.9586,
+      "step": 29560
+    },
+    {
+      "epoch": 0.6963873581084263,
+      "grad_norm": 2.5858304500579834,
+      "learning_rate": 6.078375959681598e-05,
+      "loss": 2.2042,
+      "step": 29570
+    },
+    {
+      "epoch": 0.6966228627949697,
+      "grad_norm": 2.5795745849609375,
+      "learning_rate": 6.073665865950733e-05,
+      "loss": 2.159,
+      "step": 29580
+    },
+    {
+      "epoch": 0.6968583674815129,
+      "grad_norm": 2.2104289531707764,
+      "learning_rate": 6.068955772219867e-05,
+      "loss": 1.865,
+      "step": 29590
+    },
+    {
+      "epoch": 0.6970938721680562,
+      "grad_norm": 1.7108197212219238,
+      "learning_rate": 6.064245678489002e-05,
+      "loss": 2.1158,
+      "step": 29600
+    },
+    {
+      "epoch": 0.6973293768545994,
+      "grad_norm": 2.610651731491089,
+      "learning_rate": 6.059535584758137e-05,
+      "loss": 2.0603,
+      "step": 29610
+    },
+    {
+      "epoch": 0.6975648815411427,
+      "grad_norm": 1.6342675685882568,
+      "learning_rate": 6.054825491027272e-05,
+      "loss": 1.963,
+      "step": 29620
+    },
+    {
+      "epoch": 0.6978003862276859,
+      "grad_norm": 2.5833704471588135,
+      "learning_rate": 6.050115397296407e-05,
+      "loss": 2.1255,
+      "step": 29630
+    },
+    {
+      "epoch": 0.6980358909142292,
+      "grad_norm": 2.2657909393310547,
+      "learning_rate": 6.045405303565541e-05,
+      "loss": 1.8765,
+      "step": 29640
+    },
+    {
+      "epoch": 0.6982713956007724,
+      "grad_norm": 2.124436378479004,
+      "learning_rate": 6.040695209834676e-05,
+      "loss": 2.2452,
+      "step": 29650
+    },
+    {
+      "epoch": 0.6985069002873158,
+      "grad_norm": 2.2420623302459717,
+      "learning_rate": 6.0359851161038104e-05,
+      "loss": 2.1147,
+      "step": 29660
+    },
+    {
+      "epoch": 0.698742404973859,
+      "grad_norm": 3.1197521686553955,
+      "learning_rate": 6.0312750223729455e-05,
+      "loss": 2.1481,
+      "step": 29670
+    },
+    {
+      "epoch": 0.6989779096604023,
+      "grad_norm": 2.0020577907562256,
+      "learning_rate": 6.02656492864208e-05,
+      "loss": 2.1455,
+      "step": 29680
+    },
+    {
+      "epoch": 0.6992134143469455,
+      "grad_norm": 3.46406888961792,
+      "learning_rate": 6.021854834911215e-05,
+      "loss": 1.8329,
+      "step": 29690
+    },
+    {
+      "epoch": 0.6994489190334887,
+      "grad_norm": 2.138166904449463,
+      "learning_rate": 6.017144741180349e-05,
+      "loss": 2.2264,
+      "step": 29700
+    },
+    {
+      "epoch": 0.699684423720032,
+      "grad_norm": 2.236649751663208,
+      "learning_rate": 6.012434647449484e-05,
+      "loss": 1.9857,
+      "step": 29710
+    },
+    {
+      "epoch": 0.6999199284065752,
+      "grad_norm": 2.371490955352783,
+      "learning_rate": 6.0077245537186185e-05,
+      "loss": 2.2641,
+      "step": 29720
+    },
+    {
+      "epoch": 0.7001554330931186,
+      "grad_norm": 1.9338253736495972,
+      "learning_rate": 6.0030144599877536e-05,
+      "loss": 1.7444,
+      "step": 29730
+    },
+    {
+      "epoch": 0.7003909377796618,
+      "grad_norm": 2.1143953800201416,
+      "learning_rate": 5.9983043662568886e-05,
+      "loss": 2.0956,
+      "step": 29740
+    },
+    {
+      "epoch": 0.7006264424662051,
+      "grad_norm": 2.2336223125457764,
+      "learning_rate": 5.993594272526023e-05,
+      "loss": 2.2865,
+      "step": 29750
+    },
+    {
+      "epoch": 0.7008619471527483,
+      "grad_norm": 2.3659603595733643,
+      "learning_rate": 5.9888841787951586e-05,
+      "loss": 1.9793,
+      "step": 29760
+    },
+    {
+      "epoch": 0.7010974518392916,
+      "grad_norm": 2.853128671646118,
+      "learning_rate": 5.984174085064294e-05,
+      "loss": 2.1234,
+      "step": 29770
+    },
+    {
+      "epoch": 0.7013329565258348,
+      "grad_norm": 2.3596596717834473,
+      "learning_rate": 5.979463991333428e-05,
+      "loss": 1.8743,
+      "step": 29780
+    },
+    {
+      "epoch": 0.7015684612123781,
+      "grad_norm": 1.9511758089065552,
+      "learning_rate": 5.974753897602563e-05,
+      "loss": 2.1135,
+      "step": 29790
+    },
+    {
+      "epoch": 0.7018039658989214,
+      "grad_norm": 2.2908072471618652,
+      "learning_rate": 5.9700438038716974e-05,
+      "loss": 2.2518,
+      "step": 29800
+    },
+    {
+      "epoch": 0.7020394705854647,
+      "grad_norm": 2.339449167251587,
+      "learning_rate": 5.9653337101408324e-05,
+      "loss": 2.2305,
+      "step": 29810
+    },
+    {
+      "epoch": 0.7022749752720079,
+      "grad_norm": 2.239489793777466,
+      "learning_rate": 5.9606236164099674e-05,
+      "loss": 2.0766,
+      "step": 29820
+    },
+    {
+      "epoch": 0.7025104799585512,
+      "grad_norm": 1.768304467201233,
+      "learning_rate": 5.955913522679102e-05,
+      "loss": 2.1229,
+      "step": 29830
+    },
+    {
+      "epoch": 0.7027459846450944,
+      "grad_norm": 1.8999934196472168,
+      "learning_rate": 5.951203428948237e-05,
+      "loss": 1.9637,
+      "step": 29840
+    },
+    {
+      "epoch": 0.7029814893316377,
+      "grad_norm": 2.1406350135803223,
+      "learning_rate": 5.946493335217371e-05,
+      "loss": 1.8846,
+      "step": 29850
+    },
+    {
+      "epoch": 0.7032169940181809,
+      "grad_norm": 2.1771881580352783,
+      "learning_rate": 5.941783241486506e-05,
+      "loss": 1.9483,
+      "step": 29860
+    },
+    {
+      "epoch": 0.7034524987047243,
+      "grad_norm": 1.9195053577423096,
+      "learning_rate": 5.9370731477556405e-05,
+      "loss": 1.8552,
+      "step": 29870
+    },
+    {
+      "epoch": 0.7036880033912675,
+      "grad_norm": 2.817622423171997,
+      "learning_rate": 5.9323630540247756e-05,
+      "loss": 2.0414,
+      "step": 29880
+    },
+    {
+      "epoch": 0.7039235080778108,
+      "grad_norm": 1.7897940874099731,
+      "learning_rate": 5.92765296029391e-05,
+      "loss": 1.8892,
+      "step": 29890
+    },
+    {
+      "epoch": 0.704159012764354,
+      "grad_norm": 2.211215019226074,
+      "learning_rate": 5.922942866563045e-05,
+      "loss": 2.002,
+      "step": 29900
+    },
+    {
+      "epoch": 0.7043945174508973,
+      "grad_norm": 2.0046420097351074,
+      "learning_rate": 5.918232772832179e-05,
+      "loss": 1.9973,
+      "step": 29910
+    },
+    {
+      "epoch": 0.7046300221374405,
+      "grad_norm": 2.126875638961792,
+      "learning_rate": 5.913522679101314e-05,
+      "loss": 2.044,
+      "step": 29920
+    },
+    {
+      "epoch": 0.7048655268239838,
+      "grad_norm": 2.909090995788574,
+      "learning_rate": 5.908812585370449e-05,
+      "loss": 1.7947,
+      "step": 29930
+    },
+    {
+      "epoch": 0.705101031510527,
+      "grad_norm": 2.0971028804779053,
+      "learning_rate": 5.904102491639584e-05,
+      "loss": 1.9566,
+      "step": 29940
+    },
+    {
+      "epoch": 0.7053365361970704,
+      "grad_norm": 2.6434972286224365,
+      "learning_rate": 5.899392397908719e-05,
+      "loss": 2.1791,
+      "step": 29950
+    },
+    {
+      "epoch": 0.7055720408836136,
+      "grad_norm": 1.7850927114486694,
+      "learning_rate": 5.894682304177853e-05,
+      "loss": 2.0034,
+      "step": 29960
+    },
+    {
+      "epoch": 0.7058075455701568,
+      "grad_norm": 1.7622065544128418,
+      "learning_rate": 5.889972210446988e-05,
+      "loss": 1.9323,
+      "step": 29970
+    },
+    {
+      "epoch": 0.7060430502567001,
+      "grad_norm": 2.0412206649780273,
+      "learning_rate": 5.8852621167161224e-05,
+      "loss": 2.0019,
+      "step": 29980
+    },
+    {
+      "epoch": 0.7062785549432433,
+      "grad_norm": 1.7954665422439575,
+      "learning_rate": 5.8805520229852574e-05,
+      "loss": 1.9332,
+      "step": 29990
+    },
+    {
+      "epoch": 0.7065140596297866,
+      "grad_norm": 4.530249118804932,
+      "learning_rate": 5.875841929254392e-05,
+      "loss": 2.0373,
+      "step": 30000
+    },
+    {
+      "epoch": 0.7067495643163298,
+      "grad_norm": 1.6885002851486206,
+      "learning_rate": 5.871131835523527e-05,
+      "loss": 1.9966,
+      "step": 30010
+    },
+    {
+      "epoch": 0.7069850690028732,
+      "grad_norm": 2.3572335243225098,
+      "learning_rate": 5.866421741792661e-05,
+      "loss": 1.9012,
+      "step": 30020
+    },
+    {
+      "epoch": 0.7072205736894164,
+      "grad_norm": 1.8825509548187256,
+      "learning_rate": 5.861711648061796e-05,
+      "loss": 2.2303,
+      "step": 30030
+    },
+    {
+      "epoch": 0.7074560783759597,
+      "grad_norm": 3.2787723541259766,
+      "learning_rate": 5.857001554330932e-05,
+      "loss": 2.2594,
+      "step": 30040
+    },
+    {
+      "epoch": 0.7076915830625029,
+      "grad_norm": 2.52473783493042,
+      "learning_rate": 5.852291460600067e-05,
+      "loss": 2.1244,
+      "step": 30050
+    },
+    {
+      "epoch": 0.7079270877490462,
+      "grad_norm": 2.1791837215423584,
+      "learning_rate": 5.847581366869201e-05,
+      "loss": 1.973,
+      "step": 30060
+    },
+    {
+      "epoch": 0.7081625924355894,
+      "grad_norm": 1.9965521097183228,
+      "learning_rate": 5.842871273138336e-05,
+      "loss": 1.8938,
+      "step": 30070
+    },
+    {
+      "epoch": 0.7083980971221328,
+      "grad_norm": 1.438650131225586,
+      "learning_rate": 5.8381611794074706e-05,
+      "loss": 1.9799,
+      "step": 30080
+    },
+    {
+      "epoch": 0.708633601808676,
+      "grad_norm": 1.926436185836792,
+      "learning_rate": 5.8334510856766056e-05,
+      "loss": 2.0451,
+      "step": 30090
+    },
+    {
+      "epoch": 0.7088691064952193,
+      "grad_norm": 1.62642240524292,
+      "learning_rate": 5.82874099194574e-05,
+      "loss": 2.0659,
+      "step": 30100
+    },
+    {
+      "epoch": 0.7091046111817625,
+      "grad_norm": 2.5094661712646484,
+      "learning_rate": 5.824030898214875e-05,
+      "loss": 1.9782,
+      "step": 30110
+    },
+    {
+      "epoch": 0.7093401158683058,
+      "grad_norm": 1.6265511512756348,
+      "learning_rate": 5.81932080448401e-05,
+      "loss": 1.9198,
+      "step": 30120
+    },
+    {
+      "epoch": 0.709575620554849,
+      "grad_norm": 2.2201929092407227,
+      "learning_rate": 5.8146107107531444e-05,
+      "loss": 1.962,
+      "step": 30130
+    },
+    {
+      "epoch": 0.7098111252413923,
+      "grad_norm": 2.028233051300049,
+      "learning_rate": 5.8099006170222794e-05,
+      "loss": 2.0659,
+      "step": 30140
+    },
+    {
+      "epoch": 0.7100466299279355,
+      "grad_norm": 2.96724796295166,
+      "learning_rate": 5.805190523291414e-05,
+      "loss": 2.0536,
+      "step": 30150
+    },
+    {
+      "epoch": 0.7102821346144789,
+      "grad_norm": 2.110179901123047,
+      "learning_rate": 5.800480429560549e-05,
+      "loss": 2.0094,
+      "step": 30160
+    },
+    {
+      "epoch": 0.7105176393010221,
+      "grad_norm": 2.6137187480926514,
+      "learning_rate": 5.795770335829683e-05,
+      "loss": 1.9766,
+      "step": 30170
+    },
+    {
+      "epoch": 0.7107531439875654,
+      "grad_norm": 2.006112813949585,
+      "learning_rate": 5.791060242098818e-05,
+      "loss": 2.0879,
+      "step": 30180
+    },
+    {
+      "epoch": 0.7109886486741086,
+      "grad_norm": 2.337247848510742,
+      "learning_rate": 5.7863501483679525e-05,
+      "loss": 2.038,
+      "step": 30190
+    },
+    {
+      "epoch": 0.7112241533606519,
+      "grad_norm": 1.9541200399398804,
+      "learning_rate": 5.7816400546370875e-05,
+      "loss": 1.9604,
+      "step": 30200
+    },
+    {
+      "epoch": 0.7114596580471951,
+      "grad_norm": 1.8182893991470337,
+      "learning_rate": 5.776929960906222e-05,
+      "loss": 1.8738,
+      "step": 30210
+    },
+    {
+      "epoch": 0.7116951627337385,
+      "grad_norm": 2.3723807334899902,
+      "learning_rate": 5.772219867175357e-05,
+      "loss": 2.1882,
+      "step": 30220
+    },
+    {
+      "epoch": 0.7119306674202817,
+      "grad_norm": 2.4405813217163086,
+      "learning_rate": 5.767509773444492e-05,
+      "loss": 1.9004,
+      "step": 30230
+    },
+    {
+      "epoch": 0.712166172106825,
+      "grad_norm": 2.4666452407836914,
+      "learning_rate": 5.762799679713626e-05,
+      "loss": 2.0491,
+      "step": 30240
+    },
+    {
+      "epoch": 0.7124016767933682,
+      "grad_norm": 2.4087648391723633,
+      "learning_rate": 5.758089585982761e-05,
+      "loss": 2.1029,
+      "step": 30250
+    },
+    {
+      "epoch": 0.7126371814799114,
+      "grad_norm": 2.2330007553100586,
+      "learning_rate": 5.7533794922518956e-05,
+      "loss": 1.9501,
+      "step": 30260
+    },
+    {
+      "epoch": 0.7128726861664547,
+      "grad_norm": 2.045724868774414,
+      "learning_rate": 5.748669398521031e-05,
+      "loss": 2.0743,
+      "step": 30270
+    },
+    {
+      "epoch": 0.7131081908529979,
+      "grad_norm": 2.37623929977417,
+      "learning_rate": 5.743959304790165e-05,
+      "loss": 2.086,
+      "step": 30280
+    },
+    {
+      "epoch": 0.7133436955395412,
+      "grad_norm": 2.594832181930542,
+      "learning_rate": 5.7392492110593e-05,
+      "loss": 2.0116,
+      "step": 30290
+    },
+    {
+      "epoch": 0.7135792002260845,
+      "grad_norm": 2.3839049339294434,
+      "learning_rate": 5.7345391173284344e-05,
+      "loss": 2.1416,
+      "step": 30300
+    },
+    {
+      "epoch": 0.7138147049126278,
+      "grad_norm": 2.3938348293304443,
+      "learning_rate": 5.7298290235975694e-05,
+      "loss": 2.15,
+      "step": 30310
+    },
+    {
+      "epoch": 0.714050209599171,
+      "grad_norm": 1.7177648544311523,
+      "learning_rate": 5.725118929866705e-05,
+      "loss": 1.8613,
+      "step": 30320
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 2.2762975692749023,
+      "learning_rate": 5.72040883613584e-05,
+      "loss": 1.8133,
+      "step": 30330
+    },
+    {
+      "epoch": 0.7145212189722575,
+      "grad_norm": 2.3679895401000977,
+      "learning_rate": 5.7156987424049745e-05,
+      "loss": 1.9782,
+      "step": 30340
+    },
+    {
+      "epoch": 0.7147567236588008,
+      "grad_norm": 2.099062204360962,
+      "learning_rate": 5.7109886486741095e-05,
+      "loss": 2.065,
+      "step": 30350
+    },
+    {
+      "epoch": 0.714992228345344,
+      "grad_norm": 1.8509316444396973,
+      "learning_rate": 5.706278554943244e-05,
+      "loss": 2.0064,
+      "step": 30360
+    },
+    {
+      "epoch": 0.7152277330318874,
+      "grad_norm": 2.6055731773376465,
+      "learning_rate": 5.701568461212379e-05,
+      "loss": 2.21,
+      "step": 30370
+    },
+    {
+      "epoch": 0.7154632377184306,
+      "grad_norm": 2.160001754760742,
+      "learning_rate": 5.696858367481513e-05,
+      "loss": 2.0922,
+      "step": 30380
+    },
+    {
+      "epoch": 0.7156987424049739,
+      "grad_norm": 2.1145455837249756,
+      "learning_rate": 5.692148273750648e-05,
+      "loss": 2.0778,
+      "step": 30390
+    },
+    {
+      "epoch": 0.7159342470915171,
+      "grad_norm": 2.1301040649414062,
+      "learning_rate": 5.6874381800197826e-05,
+      "loss": 1.9435,
+      "step": 30400
+    },
+    {
+      "epoch": 0.7161697517780604,
+      "grad_norm": 2.1226649284362793,
+      "learning_rate": 5.6827280862889176e-05,
+      "loss": 2.0785,
+      "step": 30410
+    },
+    {
+      "epoch": 0.7164052564646036,
+      "grad_norm": 2.950019359588623,
+      "learning_rate": 5.6780179925580526e-05,
+      "loss": 1.7644,
+      "step": 30420
+    },
+    {
+      "epoch": 0.7166407611511469,
+      "grad_norm": 2.4695847034454346,
+      "learning_rate": 5.673307898827187e-05,
+      "loss": 2.2652,
+      "step": 30430
+    },
+    {
+      "epoch": 0.7168762658376902,
+      "grad_norm": 2.0165791511535645,
+      "learning_rate": 5.668597805096322e-05,
+      "loss": 1.9458,
+      "step": 30440
+    },
+    {
+      "epoch": 0.7171117705242335,
+      "grad_norm": 2.2012386322021484,
+      "learning_rate": 5.6638877113654564e-05,
+      "loss": 1.9355,
+      "step": 30450
+    },
+    {
+      "epoch": 0.7173472752107767,
+      "grad_norm": 2.5340282917022705,
+      "learning_rate": 5.6591776176345914e-05,
+      "loss": 2.1552,
+      "step": 30460
+    },
+    {
+      "epoch": 0.71758277989732,
+      "grad_norm": 2.569276809692383,
+      "learning_rate": 5.654467523903726e-05,
+      "loss": 1.962,
+      "step": 30470
+    },
+    {
+      "epoch": 0.7178182845838632,
+      "grad_norm": 2.128749132156372,
+      "learning_rate": 5.649757430172861e-05,
+      "loss": 2.036,
+      "step": 30480
+    },
+    {
+      "epoch": 0.7180537892704065,
+      "grad_norm": 2.065203905105591,
+      "learning_rate": 5.645047336441995e-05,
+      "loss": 2.0853,
+      "step": 30490
+    },
+    {
+      "epoch": 0.7182892939569497,
+      "grad_norm": 2.0871028900146484,
+      "learning_rate": 5.64033724271113e-05,
+      "loss": 2.2986,
+      "step": 30500
+    },
+    {
+      "epoch": 0.7185247986434931,
+      "grad_norm": 1.4431718587875366,
+      "learning_rate": 5.6356271489802645e-05,
+      "loss": 1.9797,
+      "step": 30510
+    },
+    {
+      "epoch": 0.7187603033300363,
+      "grad_norm": 2.06838321685791,
+      "learning_rate": 5.6309170552493995e-05,
+      "loss": 1.945,
+      "step": 30520
+    },
+    {
+      "epoch": 0.7189958080165796,
+      "grad_norm": 1.6888794898986816,
+      "learning_rate": 5.6262069615185345e-05,
+      "loss": 1.9909,
+      "step": 30530
+    },
+    {
+      "epoch": 0.7192313127031228,
+      "grad_norm": 2.028390884399414,
+      "learning_rate": 5.621496867787669e-05,
+      "loss": 1.8028,
+      "step": 30540
+    },
+    {
+      "epoch": 0.719466817389666,
+      "grad_norm": 2.6491353511810303,
+      "learning_rate": 5.616786774056804e-05,
+      "loss": 2.1648,
+      "step": 30550
+    },
+    {
+      "epoch": 0.7197023220762093,
+      "grad_norm": 2.0216126441955566,
+      "learning_rate": 5.612076680325938e-05,
+      "loss": 1.9735,
+      "step": 30560
+    },
+    {
+      "epoch": 0.7199378267627525,
+      "grad_norm": 3.1546339988708496,
+      "learning_rate": 5.607366586595073e-05,
+      "loss": 2.1136,
+      "step": 30570
+    },
+    {
+      "epoch": 0.7201733314492959,
+      "grad_norm": 2.0467398166656494,
+      "learning_rate": 5.6026564928642076e-05,
+      "loss": 1.8179,
+      "step": 30580
+    },
+    {
+      "epoch": 0.7204088361358391,
+      "grad_norm": 1.9955779314041138,
+      "learning_rate": 5.5979463991333426e-05,
+      "loss": 1.9321,
+      "step": 30590
+    },
+    {
+      "epoch": 0.7206443408223824,
+      "grad_norm": 2.4097037315368652,
+      "learning_rate": 5.593236305402477e-05,
+      "loss": 2.0949,
+      "step": 30600
+    },
+    {
+      "epoch": 0.7208798455089256,
+      "grad_norm": 2.0927698612213135,
+      "learning_rate": 5.5885262116716134e-05,
+      "loss": 1.9115,
+      "step": 30610
+    },
+    {
+      "epoch": 0.7211153501954689,
+      "grad_norm": 2.086876630783081,
+      "learning_rate": 5.583816117940748e-05,
+      "loss": 2.3412,
+      "step": 30620
+    },
+    {
+      "epoch": 0.7213508548820121,
+      "grad_norm": 1.9654732942581177,
+      "learning_rate": 5.579106024209883e-05,
+      "loss": 2.1572,
+      "step": 30630
+    },
+    {
+      "epoch": 0.7215863595685554,
+      "grad_norm": 1.9568740129470825,
+      "learning_rate": 5.574395930479017e-05,
+      "loss": 1.8459,
+      "step": 30640
+    },
+    {
+      "epoch": 0.7218218642550986,
+      "grad_norm": 1.8348342180252075,
+      "learning_rate": 5.569685836748152e-05,
+      "loss": 1.8352,
+      "step": 30650
+    },
+    {
+      "epoch": 0.722057368941642,
+      "grad_norm": 2.3031091690063477,
+      "learning_rate": 5.5649757430172865e-05,
+      "loss": 1.9543,
+      "step": 30660
+    },
+    {
+      "epoch": 0.7222928736281852,
+      "grad_norm": 2.4159276485443115,
+      "learning_rate": 5.5602656492864215e-05,
+      "loss": 2.1472,
+      "step": 30670
+    },
+    {
+      "epoch": 0.7225283783147285,
+      "grad_norm": 2.4604334831237793,
+      "learning_rate": 5.555555555555556e-05,
+      "loss": 2.1093,
+      "step": 30680
+    },
+    {
+      "epoch": 0.7227638830012717,
+      "grad_norm": 2.7177553176879883,
+      "learning_rate": 5.550845461824691e-05,
+      "loss": 2.0667,
+      "step": 30690
+    },
+    {
+      "epoch": 0.722999387687815,
+      "grad_norm": 2.469963788986206,
+      "learning_rate": 5.546135368093825e-05,
+      "loss": 2.0233,
+      "step": 30700
+    },
+    {
+      "epoch": 0.7232348923743582,
+      "grad_norm": 2.972891330718994,
+      "learning_rate": 5.54142527436296e-05,
+      "loss": 1.9273,
+      "step": 30710
+    },
+    {
+      "epoch": 0.7234703970609015,
+      "grad_norm": 2.8329389095306396,
+      "learning_rate": 5.536715180632095e-05,
+      "loss": 2.0522,
+      "step": 30720
+    },
+    {
+      "epoch": 0.7237059017474448,
+      "grad_norm": 2.5914034843444824,
+      "learning_rate": 5.5320050869012296e-05,
+      "loss": 1.9087,
+      "step": 30730
+    },
+    {
+      "epoch": 0.7239414064339881,
+      "grad_norm": 2.375659227371216,
+      "learning_rate": 5.5272949931703646e-05,
+      "loss": 1.9656,
+      "step": 30740
+    },
+    {
+      "epoch": 0.7241769111205313,
+      "grad_norm": 2.802253246307373,
+      "learning_rate": 5.522584899439499e-05,
+      "loss": 1.9603,
+      "step": 30750
+    },
+    {
+      "epoch": 0.7244124158070746,
+      "grad_norm": 2.0423829555511475,
+      "learning_rate": 5.517874805708634e-05,
+      "loss": 2.0772,
+      "step": 30760
+    },
+    {
+      "epoch": 0.7246479204936178,
+      "grad_norm": 2.187680959701538,
+      "learning_rate": 5.513164711977768e-05,
+      "loss": 1.9225,
+      "step": 30770
+    },
+    {
+      "epoch": 0.7248834251801611,
+      "grad_norm": 2.442068576812744,
+      "learning_rate": 5.5084546182469034e-05,
+      "loss": 1.9933,
+      "step": 30780
+    },
+    {
+      "epoch": 0.7251189298667043,
+      "grad_norm": 2.3445839881896973,
+      "learning_rate": 5.503744524516038e-05,
+      "loss": 2.2684,
+      "step": 30790
+    },
+    {
+      "epoch": 0.7253544345532477,
+      "grad_norm": 2.0452582836151123,
+      "learning_rate": 5.499034430785173e-05,
+      "loss": 2.1643,
+      "step": 30800
+    },
+    {
+      "epoch": 0.7255899392397909,
+      "grad_norm": 2.025725841522217,
+      "learning_rate": 5.494324337054307e-05,
+      "loss": 1.9175,
+      "step": 30810
+    },
+    {
+      "epoch": 0.7258254439263341,
+      "grad_norm": 2.7299511432647705,
+      "learning_rate": 5.489614243323442e-05,
+      "loss": 2.0061,
+      "step": 30820
+    },
+    {
+      "epoch": 0.7260609486128774,
+      "grad_norm": 2.5968010425567627,
+      "learning_rate": 5.484904149592577e-05,
+      "loss": 1.905,
+      "step": 30830
+    },
+    {
+      "epoch": 0.7262964532994206,
+      "grad_norm": 2.217876434326172,
+      "learning_rate": 5.4801940558617115e-05,
+      "loss": 2.0114,
+      "step": 30840
+    },
+    {
+      "epoch": 0.7265319579859639,
+      "grad_norm": 3.317579984664917,
+      "learning_rate": 5.4754839621308465e-05,
+      "loss": 1.9557,
+      "step": 30850
+    },
+    {
+      "epoch": 0.7267674626725071,
+      "grad_norm": 3.0695741176605225,
+      "learning_rate": 5.470773868399981e-05,
+      "loss": 2.0315,
+      "step": 30860
+    },
+    {
+      "epoch": 0.7270029673590505,
+      "grad_norm": 2.6726536750793457,
+      "learning_rate": 5.466063774669116e-05,
+      "loss": 1.9777,
+      "step": 30870
+    },
+    {
+      "epoch": 0.7272384720455937,
+      "grad_norm": 2.8172085285186768,
+      "learning_rate": 5.46135368093825e-05,
+      "loss": 2.044,
+      "step": 30880
+    },
+    {
+      "epoch": 0.727473976732137,
+      "grad_norm": 2.0183653831481934,
+      "learning_rate": 5.456643587207386e-05,
+      "loss": 1.9781,
+      "step": 30890
+    },
+    {
+      "epoch": 0.7277094814186802,
+      "grad_norm": 2.0858523845672607,
+      "learning_rate": 5.451933493476521e-05,
+      "loss": 1.7131,
+      "step": 30900
+    },
+    {
+      "epoch": 0.7279449861052235,
+      "grad_norm": 2.115558624267578,
+      "learning_rate": 5.447223399745656e-05,
+      "loss": 1.992,
+      "step": 30910
+    },
+    {
+      "epoch": 0.7281804907917667,
+      "grad_norm": 1.972657322883606,
+      "learning_rate": 5.44251330601479e-05,
+      "loss": 1.8837,
+      "step": 30920
+    },
+    {
+      "epoch": 0.72841599547831,
+      "grad_norm": 1.986677885055542,
+      "learning_rate": 5.437803212283925e-05,
+      "loss": 2.0074,
+      "step": 30930
+    },
+    {
+      "epoch": 0.7286515001648532,
+      "grad_norm": 2.1113574504852295,
+      "learning_rate": 5.43309311855306e-05,
+      "loss": 2.0117,
+      "step": 30940
+    },
+    {
+      "epoch": 0.7288870048513966,
+      "grad_norm": 2.277940034866333,
+      "learning_rate": 5.428383024822195e-05,
+      "loss": 2.0174,
+      "step": 30950
+    },
+    {
+      "epoch": 0.7291225095379398,
+      "grad_norm": 2.1765830516815186,
+      "learning_rate": 5.423672931091329e-05,
+      "loss": 2.0942,
+      "step": 30960
+    },
+    {
+      "epoch": 0.7293580142244831,
+      "grad_norm": 2.458887815475464,
+      "learning_rate": 5.418962837360464e-05,
+      "loss": 2.0839,
+      "step": 30970
+    },
+    {
+      "epoch": 0.7295935189110263,
+      "grad_norm": 2.4584996700286865,
+      "learning_rate": 5.4142527436295984e-05,
+      "loss": 1.8682,
+      "step": 30980
+    },
+    {
+      "epoch": 0.7298290235975696,
+      "grad_norm": 2.042698860168457,
+      "learning_rate": 5.4095426498987334e-05,
+      "loss": 1.9744,
+      "step": 30990
+    },
+    {
+      "epoch": 0.7300645282841128,
+      "grad_norm": 2.2433009147644043,
+      "learning_rate": 5.404832556167868e-05,
+      "loss": 2.0678,
+      "step": 31000
+    },
+    {
+      "epoch": 0.7303000329706562,
+      "grad_norm": 3.455995559692383,
+      "learning_rate": 5.400122462437003e-05,
+      "loss": 2.0532,
+      "step": 31010
+    },
+    {
+      "epoch": 0.7305355376571994,
+      "grad_norm": 3.3080856800079346,
+      "learning_rate": 5.395412368706138e-05,
+      "loss": 2.1967,
+      "step": 31020
+    },
+    {
+      "epoch": 0.7307710423437427,
+      "grad_norm": 1.755066990852356,
+      "learning_rate": 5.390702274975272e-05,
+      "loss": 1.9065,
+      "step": 31030
+    },
+    {
+      "epoch": 0.7310065470302859,
+      "grad_norm": 2.1286559104919434,
+      "learning_rate": 5.385992181244407e-05,
+      "loss": 1.8758,
+      "step": 31040
+    },
+    {
+      "epoch": 0.7312420517168292,
+      "grad_norm": 1.8122589588165283,
+      "learning_rate": 5.3812820875135416e-05,
+      "loss": 1.9088,
+      "step": 31050
+    },
+    {
+      "epoch": 0.7314775564033724,
+      "grad_norm": 2.3942129611968994,
+      "learning_rate": 5.3765719937826766e-05,
+      "loss": 1.9794,
+      "step": 31060
+    },
+    {
+      "epoch": 0.7317130610899157,
+      "grad_norm": 2.5759265422821045,
+      "learning_rate": 5.371861900051811e-05,
+      "loss": 2.1059,
+      "step": 31070
+    },
+    {
+      "epoch": 0.731948565776459,
+      "grad_norm": 2.152968406677246,
+      "learning_rate": 5.367151806320946e-05,
+      "loss": 2.1384,
+      "step": 31080
+    },
+    {
+      "epoch": 0.7321840704630023,
+      "grad_norm": 1.7909654378890991,
+      "learning_rate": 5.36244171259008e-05,
+      "loss": 2.1023,
+      "step": 31090
+    },
+    {
+      "epoch": 0.7324195751495455,
+      "grad_norm": 2.8516340255737305,
+      "learning_rate": 5.357731618859215e-05,
+      "loss": 2.0101,
+      "step": 31100
+    },
+    {
+      "epoch": 0.7326550798360887,
+      "grad_norm": 3.418398857116699,
+      "learning_rate": 5.35302152512835e-05,
+      "loss": 2.0228,
+      "step": 31110
+    },
+    {
+      "epoch": 0.732890584522632,
+      "grad_norm": 1.9390182495117188,
+      "learning_rate": 5.348311431397485e-05,
+      "loss": 1.8994,
+      "step": 31120
+    },
+    {
+      "epoch": 0.7331260892091752,
+      "grad_norm": 2.929248809814453,
+      "learning_rate": 5.34360133766662e-05,
+      "loss": 2.017,
+      "step": 31130
+    },
+    {
+      "epoch": 0.7333615938957185,
+      "grad_norm": 2.084099531173706,
+      "learning_rate": 5.338891243935754e-05,
+      "loss": 2.0448,
+      "step": 31140
+    },
+    {
+      "epoch": 0.7335970985822617,
+      "grad_norm": 1.9358183145523071,
+      "learning_rate": 5.334181150204889e-05,
+      "loss": 1.9732,
+      "step": 31150
+    },
+    {
+      "epoch": 0.7338326032688051,
+      "grad_norm": 1.9566363096237183,
+      "learning_rate": 5.3294710564740234e-05,
+      "loss": 2.023,
+      "step": 31160
+    },
+    {
+      "epoch": 0.7340681079553483,
+      "grad_norm": 2.517813205718994,
+      "learning_rate": 5.324760962743159e-05,
+      "loss": 1.9894,
+      "step": 31170
+    },
+    {
+      "epoch": 0.7343036126418916,
+      "grad_norm": 1.871909737586975,
+      "learning_rate": 5.320050869012294e-05,
+      "loss": 2.1048,
+      "step": 31180
+    },
+    {
+      "epoch": 0.7345391173284348,
+      "grad_norm": 1.6476243734359741,
+      "learning_rate": 5.3153407752814285e-05,
+      "loss": 1.8816,
+      "step": 31190
+    },
+    {
+      "epoch": 0.7347746220149781,
+      "grad_norm": 2.7056660652160645,
+      "learning_rate": 5.3106306815505635e-05,
+      "loss": 1.8774,
+      "step": 31200
+    },
+    {
+      "epoch": 0.7350101267015213,
+      "grad_norm": 2.0394725799560547,
+      "learning_rate": 5.3059205878196986e-05,
+      "loss": 1.9307,
+      "step": 31210
+    },
+    {
+      "epoch": 0.7352456313880646,
+      "grad_norm": 2.695197105407715,
+      "learning_rate": 5.301210494088833e-05,
+      "loss": 1.842,
+      "step": 31220
+    },
+    {
+      "epoch": 0.7354811360746079,
+      "grad_norm": 2.9238481521606445,
+      "learning_rate": 5.296500400357968e-05,
+      "loss": 2.1305,
+      "step": 31230
+    },
+    {
+      "epoch": 0.7357166407611512,
+      "grad_norm": 2.0291872024536133,
+      "learning_rate": 5.291790306627102e-05,
+      "loss": 1.9462,
+      "step": 31240
+    },
+    {
+      "epoch": 0.7359521454476944,
+      "grad_norm": 2.4080393314361572,
+      "learning_rate": 5.287080212896237e-05,
+      "loss": 2.1098,
+      "step": 31250
+    },
+    {
+      "epoch": 0.7361876501342377,
+      "grad_norm": 2.3340492248535156,
+      "learning_rate": 5.2823701191653717e-05,
+      "loss": 2.135,
+      "step": 31260
+    },
+    {
+      "epoch": 0.7364231548207809,
+      "grad_norm": 2.0928733348846436,
+      "learning_rate": 5.277660025434507e-05,
+      "loss": 2.1051,
+      "step": 31270
+    },
+    {
+      "epoch": 0.7366586595073242,
+      "grad_norm": 2.603736162185669,
+      "learning_rate": 5.272949931703641e-05,
+      "loss": 2.2034,
+      "step": 31280
+    },
+    {
+      "epoch": 0.7368941641938674,
+      "grad_norm": 3.005927324295044,
+      "learning_rate": 5.268239837972776e-05,
+      "loss": 2.0459,
+      "step": 31290
+    },
+    {
+      "epoch": 0.7371296688804108,
+      "grad_norm": 2.0738449096679688,
+      "learning_rate": 5.2635297442419104e-05,
+      "loss": 2.1972,
+      "step": 31300
+    },
+    {
+      "epoch": 0.737365173566954,
+      "grad_norm": 2.542039394378662,
+      "learning_rate": 5.2588196505110454e-05,
+      "loss": 1.8309,
+      "step": 31310
+    },
+    {
+      "epoch": 0.7376006782534973,
+      "grad_norm": 2.0776071548461914,
+      "learning_rate": 5.2541095567801804e-05,
+      "loss": 2.2318,
+      "step": 31320
+    },
+    {
+      "epoch": 0.7378361829400405,
+      "grad_norm": 2.2851767539978027,
+      "learning_rate": 5.249399463049315e-05,
+      "loss": 2.1569,
+      "step": 31330
+    },
+    {
+      "epoch": 0.7380716876265838,
+      "grad_norm": 2.047200918197632,
+      "learning_rate": 5.24468936931845e-05,
+      "loss": 1.9189,
+      "step": 31340
+    },
+    {
+      "epoch": 0.738307192313127,
+      "grad_norm": 2.037782669067383,
+      "learning_rate": 5.239979275587584e-05,
+      "loss": 2.0439,
+      "step": 31350
+    },
+    {
+      "epoch": 0.7385426969996703,
+      "grad_norm": 2.144557476043701,
+      "learning_rate": 5.235269181856719e-05,
+      "loss": 1.9919,
+      "step": 31360
+    },
+    {
+      "epoch": 0.7387782016862136,
+      "grad_norm": 1.9935516119003296,
+      "learning_rate": 5.2305590881258535e-05,
+      "loss": 1.9855,
+      "step": 31370
+    },
+    {
+      "epoch": 0.7390137063727569,
+      "grad_norm": 2.8330445289611816,
+      "learning_rate": 5.2258489943949886e-05,
+      "loss": 2.1518,
+      "step": 31380
+    },
+    {
+      "epoch": 0.7392492110593001,
+      "grad_norm": 2.2874932289123535,
+      "learning_rate": 5.221138900664123e-05,
+      "loss": 2.0948,
+      "step": 31390
+    },
+    {
+      "epoch": 0.7394847157458433,
+      "grad_norm": 2.6140425205230713,
+      "learning_rate": 5.216428806933258e-05,
+      "loss": 2.0383,
+      "step": 31400
+    },
+    {
+      "epoch": 0.7397202204323866,
+      "grad_norm": 2.7529566287994385,
+      "learning_rate": 5.211718713202392e-05,
+      "loss": 2.0262,
+      "step": 31410
+    },
+    {
+      "epoch": 0.7399557251189298,
+      "grad_norm": 1.8774211406707764,
+      "learning_rate": 5.207008619471527e-05,
+      "loss": 1.7626,
+      "step": 31420
+    },
+    {
+      "epoch": 0.7401912298054731,
+      "grad_norm": 2.481081008911133,
+      "learning_rate": 5.202298525740662e-05,
+      "loss": 2.2189,
+      "step": 31430
+    },
+    {
+      "epoch": 0.7404267344920163,
+      "grad_norm": 3.3495407104492188,
+      "learning_rate": 5.197588432009797e-05,
+      "loss": 2.0925,
+      "step": 31440
+    },
+    {
+      "epoch": 0.7406622391785597,
+      "grad_norm": 2.11091685295105,
+      "learning_rate": 5.1928783382789324e-05,
+      "loss": 2.1724,
+      "step": 31450
+    },
+    {
+      "epoch": 0.7408977438651029,
+      "grad_norm": 2.0129551887512207,
+      "learning_rate": 5.1881682445480674e-05,
+      "loss": 2.1202,
+      "step": 31460
+    },
+    {
+      "epoch": 0.7411332485516462,
+      "grad_norm": 1.9652073383331299,
+      "learning_rate": 5.183458150817202e-05,
+      "loss": 1.9064,
+      "step": 31470
+    },
+    {
+      "epoch": 0.7413687532381894,
+      "grad_norm": 2.4702093601226807,
+      "learning_rate": 5.178748057086337e-05,
+      "loss": 2.042,
+      "step": 31480
+    },
+    {
+      "epoch": 0.7416042579247327,
+      "grad_norm": 1.9683281183242798,
+      "learning_rate": 5.174037963355471e-05,
+      "loss": 1.8423,
+      "step": 31490
+    },
+    {
+      "epoch": 0.7418397626112759,
+      "grad_norm": 1.8819502592086792,
+      "learning_rate": 5.169327869624606e-05,
+      "loss": 1.9363,
+      "step": 31500
+    },
+    {
+      "epoch": 0.7420752672978193,
+      "grad_norm": 2.266874074935913,
+      "learning_rate": 5.164617775893741e-05,
+      "loss": 1.9024,
+      "step": 31510
+    },
+    {
+      "epoch": 0.7423107719843625,
+      "grad_norm": 2.113044023513794,
+      "learning_rate": 5.1599076821628755e-05,
+      "loss": 1.9809,
+      "step": 31520
+    },
+    {
+      "epoch": 0.7425462766709058,
+      "grad_norm": 2.2748920917510986,
+      "learning_rate": 5.1551975884320105e-05,
+      "loss": 2.0686,
+      "step": 31530
+    },
+    {
+      "epoch": 0.742781781357449,
+      "grad_norm": 2.078479051589966,
+      "learning_rate": 5.150487494701145e-05,
+      "loss": 2.1953,
+      "step": 31540
+    },
+    {
+      "epoch": 0.7430172860439923,
+      "grad_norm": 2.49847412109375,
+      "learning_rate": 5.14577740097028e-05,
+      "loss": 2.0343,
+      "step": 31550
+    },
+    {
+      "epoch": 0.7432527907305355,
+      "grad_norm": 2.6483688354492188,
+      "learning_rate": 5.141067307239414e-05,
+      "loss": 1.9767,
+      "step": 31560
+    },
+    {
+      "epoch": 0.7434882954170788,
+      "grad_norm": 2.081569194793701,
+      "learning_rate": 5.136357213508549e-05,
+      "loss": 2.0236,
+      "step": 31570
+    },
+    {
+      "epoch": 0.743723800103622,
+      "grad_norm": 2.007598876953125,
+      "learning_rate": 5.1316471197776836e-05,
+      "loss": 2.079,
+      "step": 31580
+    },
+    {
+      "epoch": 0.7439593047901654,
+      "grad_norm": 2.3295485973358154,
+      "learning_rate": 5.1269370260468186e-05,
+      "loss": 1.9723,
+      "step": 31590
+    },
+    {
+      "epoch": 0.7441948094767086,
+      "grad_norm": 2.001807689666748,
+      "learning_rate": 5.122226932315953e-05,
+      "loss": 2.0025,
+      "step": 31600
+    },
+    {
+      "epoch": 0.7444303141632519,
+      "grad_norm": 2.1212780475616455,
+      "learning_rate": 5.117516838585088e-05,
+      "loss": 2.1097,
+      "step": 31610
+    },
+    {
+      "epoch": 0.7446658188497951,
+      "grad_norm": 1.8547629117965698,
+      "learning_rate": 5.112806744854223e-05,
+      "loss": 1.9761,
+      "step": 31620
+    },
+    {
+      "epoch": 0.7449013235363384,
+      "grad_norm": 1.8750020265579224,
+      "learning_rate": 5.1080966511233574e-05,
+      "loss": 2.0775,
+      "step": 31630
+    },
+    {
+      "epoch": 0.7451368282228816,
+      "grad_norm": 2.390493869781494,
+      "learning_rate": 5.1033865573924924e-05,
+      "loss": 2.0353,
+      "step": 31640
+    },
+    {
+      "epoch": 0.745372332909425,
+      "grad_norm": 1.8279821872711182,
+      "learning_rate": 5.098676463661627e-05,
+      "loss": 1.958,
+      "step": 31650
+    },
+    {
+      "epoch": 0.7456078375959682,
+      "grad_norm": 2.5792455673217773,
+      "learning_rate": 5.093966369930762e-05,
+      "loss": 2.2457,
+      "step": 31660
+    },
+    {
+      "epoch": 0.7458433422825114,
+      "grad_norm": 2.464423894882202,
+      "learning_rate": 5.089256276199896e-05,
+      "loss": 1.8873,
+      "step": 31670
+    },
+    {
+      "epoch": 0.7460788469690547,
+      "grad_norm": 1.911726951599121,
+      "learning_rate": 5.084546182469031e-05,
+      "loss": 2.0966,
+      "step": 31680
+    },
+    {
+      "epoch": 0.7463143516555979,
+      "grad_norm": 2.0924248695373535,
+      "learning_rate": 5.0798360887381655e-05,
+      "loss": 2.1157,
+      "step": 31690
+    },
+    {
+      "epoch": 0.7465498563421412,
+      "grad_norm": 2.2013473510742188,
+      "learning_rate": 5.0751259950073005e-05,
+      "loss": 2.05,
+      "step": 31700
+    },
+    {
+      "epoch": 0.7467853610286844,
+      "grad_norm": 2.036799430847168,
+      "learning_rate": 5.070415901276435e-05,
+      "loss": 2.1727,
+      "step": 31710
+    },
+    {
+      "epoch": 0.7470208657152277,
+      "grad_norm": 1.8770381212234497,
+      "learning_rate": 5.06570580754557e-05,
+      "loss": 2.0502,
+      "step": 31720
+    },
+    {
+      "epoch": 0.747256370401771,
+      "grad_norm": 1.9372336864471436,
+      "learning_rate": 5.060995713814705e-05,
+      "loss": 2.1371,
+      "step": 31730
+    },
+    {
+      "epoch": 0.7474918750883143,
+      "grad_norm": 2.189121961593628,
+      "learning_rate": 5.0562856200838406e-05,
+      "loss": 2.0499,
+      "step": 31740
+    },
+    {
+      "epoch": 0.7477273797748575,
+      "grad_norm": 2.1549293994903564,
+      "learning_rate": 5.051575526352975e-05,
+      "loss": 1.9287,
+      "step": 31750
+    },
+    {
+      "epoch": 0.7479628844614008,
+      "grad_norm": 2.0482614040374756,
+      "learning_rate": 5.04686543262211e-05,
+      "loss": 2.0564,
+      "step": 31760
+    },
+    {
+      "epoch": 0.748198389147944,
+      "grad_norm": 2.111997604370117,
+      "learning_rate": 5.0421553388912443e-05,
+      "loss": 2.176,
+      "step": 31770
+    },
+    {
+      "epoch": 0.7484338938344873,
+      "grad_norm": 2.4631214141845703,
+      "learning_rate": 5.0374452451603794e-05,
+      "loss": 1.9805,
+      "step": 31780
+    },
+    {
+      "epoch": 0.7486693985210305,
+      "grad_norm": 3.002066135406494,
+      "learning_rate": 5.032735151429514e-05,
+      "loss": 2.091,
+      "step": 31790
+    },
+    {
+      "epoch": 0.7489049032075739,
+      "grad_norm": 2.605787992477417,
+      "learning_rate": 5.028025057698649e-05,
+      "loss": 1.8293,
+      "step": 31800
+    },
+    {
+      "epoch": 0.7491404078941171,
+      "grad_norm": 1.7942407131195068,
+      "learning_rate": 5.023314963967784e-05,
+      "loss": 2.0559,
+      "step": 31810
+    },
+    {
+      "epoch": 0.7493759125806604,
+      "grad_norm": 2.3381400108337402,
+      "learning_rate": 5.018604870236918e-05,
+      "loss": 2.1534,
+      "step": 31820
+    },
+    {
+      "epoch": 0.7496114172672036,
+      "grad_norm": 2.1282408237457275,
+      "learning_rate": 5.013894776506053e-05,
+      "loss": 1.6971,
+      "step": 31830
+    },
+    {
+      "epoch": 0.7498469219537469,
+      "grad_norm": 2.5377397537231445,
+      "learning_rate": 5.0091846827751875e-05,
+      "loss": 2.1498,
+      "step": 31840
+    },
+    {
+      "epoch": 0.7500824266402901,
+      "grad_norm": 2.1873764991760254,
+      "learning_rate": 5.0044745890443225e-05,
+      "loss": 1.9524,
+      "step": 31850
+    },
+    {
+      "epoch": 0.7503179313268334,
+      "grad_norm": 1.9060252904891968,
+      "learning_rate": 4.999764495313457e-05,
+      "loss": 1.9569,
+      "step": 31860
+    },
+    {
+      "epoch": 0.7505534360133767,
+      "grad_norm": 1.914854645729065,
+      "learning_rate": 4.995054401582592e-05,
+      "loss": 1.9475,
+      "step": 31870
+    },
+    {
+      "epoch": 0.75078894069992,
+      "grad_norm": 3.2591161727905273,
+      "learning_rate": 4.990344307851726e-05,
+      "loss": 2.0404,
+      "step": 31880
+    },
+    {
+      "epoch": 0.7510244453864632,
+      "grad_norm": 1.9363534450531006,
+      "learning_rate": 4.985634214120861e-05,
+      "loss": 1.9256,
+      "step": 31890
+    },
+    {
+      "epoch": 0.7512599500730065,
+      "grad_norm": 2.1461710929870605,
+      "learning_rate": 4.9809241203899956e-05,
+      "loss": 1.9352,
+      "step": 31900
+    },
+    {
+      "epoch": 0.7514954547595497,
+      "grad_norm": 2.079094648361206,
+      "learning_rate": 4.9762140266591306e-05,
+      "loss": 1.9824,
+      "step": 31910
+    },
+    {
+      "epoch": 0.751730959446093,
+      "grad_norm": 3.0540053844451904,
+      "learning_rate": 4.9715039329282656e-05,
+      "loss": 2.2878,
+      "step": 31920
+    },
+    {
+      "epoch": 0.7519664641326362,
+      "grad_norm": 2.7210094928741455,
+      "learning_rate": 4.9667938391974e-05,
+      "loss": 1.8646,
+      "step": 31930
+    },
+    {
+      "epoch": 0.7522019688191796,
+      "grad_norm": 2.256350517272949,
+      "learning_rate": 4.962083745466535e-05,
+      "loss": 1.6932,
+      "step": 31940
+    },
+    {
+      "epoch": 0.7524374735057228,
+      "grad_norm": 1.946227788925171,
+      "learning_rate": 4.95737365173567e-05,
+      "loss": 2.0174,
+      "step": 31950
+    },
+    {
+      "epoch": 0.752672978192266,
+      "grad_norm": 2.2353408336639404,
+      "learning_rate": 4.952663558004805e-05,
+      "loss": 2.0512,
+      "step": 31960
+    },
+    {
+      "epoch": 0.7529084828788093,
+      "grad_norm": 2.020702362060547,
+      "learning_rate": 4.9479534642739394e-05,
+      "loss": 2.1851,
+      "step": 31970
+    },
+    {
+      "epoch": 0.7531439875653525,
+      "grad_norm": 2.0567731857299805,
+      "learning_rate": 4.9432433705430744e-05,
+      "loss": 2.0855,
+      "step": 31980
+    },
+    {
+      "epoch": 0.7533794922518958,
+      "grad_norm": 2.680443525314331,
+      "learning_rate": 4.938533276812209e-05,
+      "loss": 2.0761,
+      "step": 31990
+    },
+    {
+      "epoch": 0.753614996938439,
+      "grad_norm": 2.444499969482422,
+      "learning_rate": 4.933823183081344e-05,
+      "loss": 1.9018,
+      "step": 32000
+    },
+    {
+      "epoch": 0.7538505016249823,
+      "grad_norm": 2.7612671852111816,
+      "learning_rate": 4.929113089350478e-05,
+      "loss": 1.9357,
+      "step": 32010
+    },
+    {
+      "epoch": 0.7540860063115256,
+      "grad_norm": 2.018686532974243,
+      "learning_rate": 4.924402995619613e-05,
+      "loss": 1.9346,
+      "step": 32020
+    },
+    {
+      "epoch": 0.7543215109980689,
+      "grad_norm": 2.3052027225494385,
+      "learning_rate": 4.9196929018887475e-05,
+      "loss": 1.9173,
+      "step": 32030
+    },
+    {
+      "epoch": 0.7545570156846121,
+      "grad_norm": 2.0256094932556152,
+      "learning_rate": 4.9149828081578826e-05,
+      "loss": 1.9119,
+      "step": 32040
+    },
+    {
+      "epoch": 0.7547925203711554,
+      "grad_norm": 2.7691421508789062,
+      "learning_rate": 4.910272714427017e-05,
+      "loss": 1.9702,
+      "step": 32050
+    },
+    {
+      "epoch": 0.7550280250576986,
+      "grad_norm": 2.235715866088867,
+      "learning_rate": 4.905562620696152e-05,
+      "loss": 1.9636,
+      "step": 32060
+    },
+    {
+      "epoch": 0.7552635297442419,
+      "grad_norm": 2.506202459335327,
+      "learning_rate": 4.900852526965287e-05,
+      "loss": 1.9745,
+      "step": 32070
+    },
+    {
+      "epoch": 0.7554990344307851,
+      "grad_norm": 2.220933198928833,
+      "learning_rate": 4.896142433234421e-05,
+      "loss": 2.2018,
+      "step": 32080
+    },
+    {
+      "epoch": 0.7557345391173285,
+      "grad_norm": 2.0806100368499756,
+      "learning_rate": 4.891432339503556e-05,
+      "loss": 2.0028,
+      "step": 32090
+    },
+    {
+      "epoch": 0.7559700438038717,
+      "grad_norm": 2.027374744415283,
+      "learning_rate": 4.8867222457726913e-05,
+      "loss": 2.0422,
+      "step": 32100
+    },
+    {
+      "epoch": 0.756205548490415,
+      "grad_norm": 2.054565191268921,
+      "learning_rate": 4.8820121520418264e-05,
+      "loss": 1.9676,
+      "step": 32110
+    },
+    {
+      "epoch": 0.7564410531769582,
+      "grad_norm": 2.5844099521636963,
+      "learning_rate": 4.877302058310961e-05,
+      "loss": 2.1372,
+      "step": 32120
+    },
+    {
+      "epoch": 0.7566765578635015,
+      "grad_norm": 2.344665765762329,
+      "learning_rate": 4.872591964580096e-05,
+      "loss": 2.2227,
+      "step": 32130
+    },
+    {
+      "epoch": 0.7569120625500447,
+      "grad_norm": 2.02775502204895,
+      "learning_rate": 4.86788187084923e-05,
+      "loss": 1.8766,
+      "step": 32140
+    },
+    {
+      "epoch": 0.757147567236588,
+      "grad_norm": 1.9430595636367798,
+      "learning_rate": 4.863171777118365e-05,
+      "loss": 1.9213,
+      "step": 32150
+    },
+    {
+      "epoch": 0.7573830719231313,
+      "grad_norm": 1.9600436687469482,
+      "learning_rate": 4.8584616833874995e-05,
+      "loss": 1.7373,
+      "step": 32160
+    },
+    {
+      "epoch": 0.7576185766096746,
+      "grad_norm": 2.5143792629241943,
+      "learning_rate": 4.8537515896566345e-05,
+      "loss": 2.1279,
+      "step": 32170
+    },
+    {
+      "epoch": 0.7578540812962178,
+      "grad_norm": 1.9432998895645142,
+      "learning_rate": 4.849041495925769e-05,
+      "loss": 2.0553,
+      "step": 32180
+    },
+    {
+      "epoch": 0.7580895859827611,
+      "grad_norm": 2.147117853164673,
+      "learning_rate": 4.844331402194904e-05,
+      "loss": 1.9338,
+      "step": 32190
+    },
+    {
+      "epoch": 0.7583250906693043,
+      "grad_norm": 1.7949907779693604,
+      "learning_rate": 4.839621308464038e-05,
+      "loss": 2.0411,
+      "step": 32200
+    },
+    {
+      "epoch": 0.7585605953558476,
+      "grad_norm": 1.945996880531311,
+      "learning_rate": 4.834911214733173e-05,
+      "loss": 1.9042,
+      "step": 32210
+    },
+    {
+      "epoch": 0.7587961000423908,
+      "grad_norm": 3.3498096466064453,
+      "learning_rate": 4.830201121002308e-05,
+      "loss": 2.2681,
+      "step": 32220
+    },
+    {
+      "epoch": 0.759031604728934,
+      "grad_norm": 1.85516357421875,
+      "learning_rate": 4.825491027271443e-05,
+      "loss": 1.9829,
+      "step": 32230
+    },
+    {
+      "epoch": 0.7592671094154774,
+      "grad_norm": 1.8677126169204712,
+      "learning_rate": 4.8207809335405776e-05,
+      "loss": 1.8494,
+      "step": 32240
+    },
+    {
+      "epoch": 0.7595026141020206,
+      "grad_norm": 1.955237865447998,
+      "learning_rate": 4.8160708398097126e-05,
+      "loss": 1.9311,
+      "step": 32250
+    },
+    {
+      "epoch": 0.7597381187885639,
+      "grad_norm": 2.534681558609009,
+      "learning_rate": 4.811360746078848e-05,
+      "loss": 2.2401,
+      "step": 32260
+    },
+    {
+      "epoch": 0.7599736234751071,
+      "grad_norm": 1.8097875118255615,
+      "learning_rate": 4.806650652347982e-05,
+      "loss": 1.8716,
+      "step": 32270
+    },
+    {
+      "epoch": 0.7602091281616504,
+      "grad_norm": 2.7228007316589355,
+      "learning_rate": 4.801940558617117e-05,
+      "loss": 1.9849,
+      "step": 32280
+    },
+    {
+      "epoch": 0.7604446328481936,
+      "grad_norm": 2.104480028152466,
+      "learning_rate": 4.7972304648862514e-05,
+      "loss": 2.1219,
+      "step": 32290
+    },
+    {
+      "epoch": 0.760680137534737,
+      "grad_norm": 2.193178176879883,
+      "learning_rate": 4.7925203711553864e-05,
+      "loss": 1.8933,
+      "step": 32300
+    },
+    {
+      "epoch": 0.7609156422212802,
+      "grad_norm": 2.3160879611968994,
+      "learning_rate": 4.787810277424521e-05,
+      "loss": 2.0682,
+      "step": 32310
+    },
+    {
+      "epoch": 0.7611511469078235,
+      "grad_norm": 1.9633328914642334,
+      "learning_rate": 4.783100183693656e-05,
+      "loss": 1.893,
+      "step": 32320
+    },
+    {
+      "epoch": 0.7613866515943667,
+      "grad_norm": 2.395254611968994,
+      "learning_rate": 4.77839008996279e-05,
+      "loss": 1.9148,
+      "step": 32330
+    },
+    {
+      "epoch": 0.76162215628091,
+      "grad_norm": 2.234740972518921,
+      "learning_rate": 4.773679996231925e-05,
+      "loss": 1.9468,
+      "step": 32340
+    },
+    {
+      "epoch": 0.7618576609674532,
+      "grad_norm": 2.9619855880737305,
+      "learning_rate": 4.7689699025010595e-05,
+      "loss": 1.7942,
+      "step": 32350
+    },
+    {
+      "epoch": 0.7620931656539965,
+      "grad_norm": 2.4581329822540283,
+      "learning_rate": 4.7642598087701945e-05,
+      "loss": 1.9489,
+      "step": 32360
+    },
+    {
+      "epoch": 0.7623286703405397,
+      "grad_norm": 2.132481098175049,
+      "learning_rate": 4.7595497150393295e-05,
+      "loss": 1.8993,
+      "step": 32370
+    },
+    {
+      "epoch": 0.7625641750270831,
+      "grad_norm": 2.9130518436431885,
+      "learning_rate": 4.7548396213084646e-05,
+      "loss": 2.1784,
+      "step": 32380
+    },
+    {
+      "epoch": 0.7627996797136263,
+      "grad_norm": 2.35318922996521,
+      "learning_rate": 4.750129527577599e-05,
+      "loss": 2.0111,
+      "step": 32390
+    },
+    {
+      "epoch": 0.7630351844001696,
+      "grad_norm": 2.9324629306793213,
+      "learning_rate": 4.745419433846734e-05,
+      "loss": 1.9419,
+      "step": 32400
+    },
+    {
+      "epoch": 0.7632706890867128,
+      "grad_norm": 1.9878653287887573,
+      "learning_rate": 4.740709340115869e-05,
+      "loss": 1.9623,
+      "step": 32410
+    },
+    {
+      "epoch": 0.7635061937732561,
+      "grad_norm": 1.8808915615081787,
+      "learning_rate": 4.735999246385003e-05,
+      "loss": 2.12,
+      "step": 32420
+    },
+    {
+      "epoch": 0.7637416984597993,
+      "grad_norm": 1.9779167175292969,
+      "learning_rate": 4.7312891526541383e-05,
+      "loss": 1.775,
+      "step": 32430
+    },
+    {
+      "epoch": 0.7639772031463427,
+      "grad_norm": 2.064002752304077,
+      "learning_rate": 4.726579058923273e-05,
+      "loss": 2.0461,
+      "step": 32440
+    },
+    {
+      "epoch": 0.7642127078328859,
+      "grad_norm": 1.897849440574646,
+      "learning_rate": 4.721868965192408e-05,
+      "loss": 1.8057,
+      "step": 32450
+    },
+    {
+      "epoch": 0.7644482125194292,
+      "grad_norm": 2.096017837524414,
+      "learning_rate": 4.717158871461542e-05,
+      "loss": 1.9373,
+      "step": 32460
+    },
+    {
+      "epoch": 0.7646837172059724,
+      "grad_norm": 2.7307066917419434,
+      "learning_rate": 4.712448777730677e-05,
+      "loss": 2.1702,
+      "step": 32470
+    },
+    {
+      "epoch": 0.7649192218925157,
+      "grad_norm": 1.8949542045593262,
+      "learning_rate": 4.7077386839998114e-05,
+      "loss": 1.9492,
+      "step": 32480
+    },
+    {
+      "epoch": 0.7651547265790589,
+      "grad_norm": 2.4673449993133545,
+      "learning_rate": 4.7030285902689465e-05,
+      "loss": 2.3952,
+      "step": 32490
+    },
+    {
+      "epoch": 0.7653902312656022,
+      "grad_norm": 2.2674577236175537,
+      "learning_rate": 4.698318496538081e-05,
+      "loss": 1.9935,
+      "step": 32500
+    },
+    {
+      "epoch": 0.7656257359521454,
+      "grad_norm": 2.0791029930114746,
+      "learning_rate": 4.6936084028072165e-05,
+      "loss": 2.0564,
+      "step": 32510
+    },
+    {
+      "epoch": 0.7658612406386887,
+      "grad_norm": 2.5843989849090576,
+      "learning_rate": 4.688898309076351e-05,
+      "loss": 1.7758,
+      "step": 32520
+    },
+    {
+      "epoch": 0.766096745325232,
+      "grad_norm": 2.141845226287842,
+      "learning_rate": 4.684188215345486e-05,
+      "loss": 1.9268,
+      "step": 32530
+    },
+    {
+      "epoch": 0.7663322500117752,
+      "grad_norm": 2.0539655685424805,
+      "learning_rate": 4.67947812161462e-05,
+      "loss": 1.9727,
+      "step": 32540
+    },
+    {
+      "epoch": 0.7665677546983185,
+      "grad_norm": 2.356175184249878,
+      "learning_rate": 4.674768027883755e-05,
+      "loss": 2.1268,
+      "step": 32550
+    },
+    {
+      "epoch": 0.7668032593848617,
+      "grad_norm": 2.6625282764434814,
+      "learning_rate": 4.67005793415289e-05,
+      "loss": 2.0887,
+      "step": 32560
+    },
+    {
+      "epoch": 0.767038764071405,
+      "grad_norm": 2.2994375228881836,
+      "learning_rate": 4.6653478404220246e-05,
+      "loss": 1.7722,
+      "step": 32570
+    },
+    {
+      "epoch": 0.7672742687579482,
+      "grad_norm": 2.1485538482666016,
+      "learning_rate": 4.6606377466911596e-05,
+      "loss": 1.914,
+      "step": 32580
+    },
+    {
+      "epoch": 0.7675097734444916,
+      "grad_norm": 3.24705171585083,
+      "learning_rate": 4.655927652960294e-05,
+      "loss": 1.9643,
+      "step": 32590
+    },
+    {
+      "epoch": 0.7677452781310348,
+      "grad_norm": 2.247652292251587,
+      "learning_rate": 4.651217559229429e-05,
+      "loss": 2.2326,
+      "step": 32600
+    },
+    {
+      "epoch": 0.7679807828175781,
+      "grad_norm": 2.2194790840148926,
+      "learning_rate": 4.6465074654985634e-05,
+      "loss": 2.0224,
+      "step": 32610
+    },
+    {
+      "epoch": 0.7682162875041213,
+      "grad_norm": 2.246136426925659,
+      "learning_rate": 4.6417973717676984e-05,
+      "loss": 2.1937,
+      "step": 32620
+    },
+    {
+      "epoch": 0.7684517921906646,
+      "grad_norm": 1.978644847869873,
+      "learning_rate": 4.637087278036833e-05,
+      "loss": 2.0933,
+      "step": 32630
+    },
+    {
+      "epoch": 0.7686872968772078,
+      "grad_norm": 2.822878360748291,
+      "learning_rate": 4.632377184305968e-05,
+      "loss": 2.1852,
+      "step": 32640
+    },
+    {
+      "epoch": 0.7689228015637511,
+      "grad_norm": 1.890211820602417,
+      "learning_rate": 4.627667090575102e-05,
+      "loss": 1.9076,
+      "step": 32650
+    },
+    {
+      "epoch": 0.7691583062502944,
+      "grad_norm": 2.3078582286834717,
+      "learning_rate": 4.622956996844238e-05,
+      "loss": 1.9756,
+      "step": 32660
+    },
+    {
+      "epoch": 0.7693938109368377,
+      "grad_norm": 2.1552579402923584,
+      "learning_rate": 4.618246903113372e-05,
+      "loss": 2.0046,
+      "step": 32670
+    },
+    {
+      "epoch": 0.7696293156233809,
+      "grad_norm": 2.208951711654663,
+      "learning_rate": 4.613536809382507e-05,
+      "loss": 2.1688,
+      "step": 32680
+    },
+    {
+      "epoch": 0.7698648203099242,
+      "grad_norm": 2.6358840465545654,
+      "learning_rate": 4.6088267156516415e-05,
+      "loss": 2.0845,
+      "step": 32690
+    },
+    {
+      "epoch": 0.7701003249964674,
+      "grad_norm": 2.166287422180176,
+      "learning_rate": 4.6041166219207765e-05,
+      "loss": 2.0353,
+      "step": 32700
+    },
+    {
+      "epoch": 0.7703358296830107,
+      "grad_norm": 2.0117709636688232,
+      "learning_rate": 4.5994065281899116e-05,
+      "loss": 2.0734,
+      "step": 32710
+    },
+    {
+      "epoch": 0.7705713343695539,
+      "grad_norm": 2.068957805633545,
+      "learning_rate": 4.594696434459046e-05,
+      "loss": 2.0584,
+      "step": 32720
+    },
+    {
+      "epoch": 0.7708068390560973,
+      "grad_norm": 2.884422540664673,
+      "learning_rate": 4.589986340728181e-05,
+      "loss": 2.0943,
+      "step": 32730
+    },
+    {
+      "epoch": 0.7710423437426405,
+      "grad_norm": 2.284824848175049,
+      "learning_rate": 4.585276246997315e-05,
+      "loss": 2.0243,
+      "step": 32740
+    },
+    {
+      "epoch": 0.7712778484291838,
+      "grad_norm": 2.4009203910827637,
+      "learning_rate": 4.58056615326645e-05,
+      "loss": 2.0731,
+      "step": 32750
+    },
+    {
+      "epoch": 0.771513353115727,
+      "grad_norm": 2.728984832763672,
+      "learning_rate": 4.5758560595355847e-05,
+      "loss": 2.1744,
+      "step": 32760
+    },
+    {
+      "epoch": 0.7717488578022703,
+      "grad_norm": 2.1261510848999023,
+      "learning_rate": 4.57114596580472e-05,
+      "loss": 2.0703,
+      "step": 32770
+    },
+    {
+      "epoch": 0.7719843624888135,
+      "grad_norm": 2.4731087684631348,
+      "learning_rate": 4.566435872073854e-05,
+      "loss": 1.8765,
+      "step": 32780
+    },
+    {
+      "epoch": 0.7722198671753568,
+      "grad_norm": 2.0476648807525635,
+      "learning_rate": 4.561725778342989e-05,
+      "loss": 2.0803,
+      "step": 32790
+    },
+    {
+      "epoch": 0.7724553718619,
+      "grad_norm": 2.126819372177124,
+      "learning_rate": 4.557015684612124e-05,
+      "loss": 1.9244,
+      "step": 32800
+    },
+    {
+      "epoch": 0.7726908765484433,
+      "grad_norm": 1.6341440677642822,
+      "learning_rate": 4.552305590881259e-05,
+      "loss": 2.0313,
+      "step": 32810
+    },
+    {
+      "epoch": 0.7729263812349866,
+      "grad_norm": 2.055776596069336,
+      "learning_rate": 4.5475954971503935e-05,
+      "loss": 2.0049,
+      "step": 32820
+    },
+    {
+      "epoch": 0.7731618859215298,
+      "grad_norm": 2.976670265197754,
+      "learning_rate": 4.5428854034195285e-05,
+      "loss": 2.3065,
+      "step": 32830
+    },
+    {
+      "epoch": 0.7733973906080731,
+      "grad_norm": 2.6509125232696533,
+      "learning_rate": 4.538175309688663e-05,
+      "loss": 2.1162,
+      "step": 32840
+    },
+    {
+      "epoch": 0.7736328952946163,
+      "grad_norm": 2.617938756942749,
+      "learning_rate": 4.533465215957798e-05,
+      "loss": 2.0919,
+      "step": 32850
+    },
+    {
+      "epoch": 0.7738683999811596,
+      "grad_norm": 2.4530789852142334,
+      "learning_rate": 4.528755122226933e-05,
+      "loss": 2.009,
+      "step": 32860
+    },
+    {
+      "epoch": 0.7741039046677028,
+      "grad_norm": 2.889514207839966,
+      "learning_rate": 4.524045028496067e-05,
+      "loss": 1.8075,
+      "step": 32870
+    },
+    {
+      "epoch": 0.7743394093542462,
+      "grad_norm": 2.127507448196411,
+      "learning_rate": 4.519334934765202e-05,
+      "loss": 2.0197,
+      "step": 32880
+    },
+    {
+      "epoch": 0.7745749140407894,
+      "grad_norm": 2.079528570175171,
+      "learning_rate": 4.5146248410343366e-05,
+      "loss": 2.0391,
+      "step": 32890
+    },
+    {
+      "epoch": 0.7748104187273327,
+      "grad_norm": 2.2359161376953125,
+      "learning_rate": 4.5099147473034716e-05,
+      "loss": 2.0698,
+      "step": 32900
+    },
+    {
+      "epoch": 0.7750459234138759,
+      "grad_norm": 1.9062904119491577,
+      "learning_rate": 4.505204653572606e-05,
+      "loss": 2.1326,
+      "step": 32910
+    },
+    {
+      "epoch": 0.7752814281004192,
+      "grad_norm": 2.303953170776367,
+      "learning_rate": 4.500494559841741e-05,
+      "loss": 2.0753,
+      "step": 32920
+    },
+    {
+      "epoch": 0.7755169327869624,
+      "grad_norm": 1.8101952075958252,
+      "learning_rate": 4.495784466110875e-05,
+      "loss": 1.8815,
+      "step": 32930
+    },
+    {
+      "epoch": 0.7757524374735058,
+      "grad_norm": 3.982452630996704,
+      "learning_rate": 4.491074372380011e-05,
+      "loss": 1.9328,
+      "step": 32940
+    },
+    {
+      "epoch": 0.775987942160049,
+      "grad_norm": 2.1649701595306396,
+      "learning_rate": 4.4863642786491454e-05,
+      "loss": 1.8076,
+      "step": 32950
+    },
+    {
+      "epoch": 0.7762234468465923,
+      "grad_norm": 1.8240478038787842,
+      "learning_rate": 4.4816541849182804e-05,
+      "loss": 2.0128,
+      "step": 32960
+    },
+    {
+      "epoch": 0.7764589515331355,
+      "grad_norm": 1.8030163049697876,
+      "learning_rate": 4.476944091187415e-05,
+      "loss": 2.0234,
+      "step": 32970
+    },
+    {
+      "epoch": 0.7766944562196788,
+      "grad_norm": 1.9468859434127808,
+      "learning_rate": 4.47223399745655e-05,
+      "loss": 1.9977,
+      "step": 32980
+    },
+    {
+      "epoch": 0.776929960906222,
+      "grad_norm": 2.0538527965545654,
+      "learning_rate": 4.467523903725684e-05,
+      "loss": 2.141,
+      "step": 32990
+    },
+    {
+      "epoch": 0.7771654655927653,
+      "grad_norm": 2.1472928524017334,
+      "learning_rate": 4.462813809994819e-05,
+      "loss": 2.1504,
+      "step": 33000
+    },
+    {
+      "epoch": 0.7774009702793085,
+      "grad_norm": 2.2194812297821045,
+      "learning_rate": 4.458103716263954e-05,
+      "loss": 2.0039,
+      "step": 33010
+    },
+    {
+      "epoch": 0.7776364749658519,
+      "grad_norm": 4.218422889709473,
+      "learning_rate": 4.4533936225330885e-05,
+      "loss": 1.922,
+      "step": 33020
+    },
+    {
+      "epoch": 0.7778719796523951,
+      "grad_norm": 2.516357421875,
+      "learning_rate": 4.4486835288022235e-05,
+      "loss": 2.0226,
+      "step": 33030
+    },
+    {
+      "epoch": 0.7781074843389384,
+      "grad_norm": 2.656029224395752,
+      "learning_rate": 4.443973435071358e-05,
+      "loss": 1.9103,
+      "step": 33040
+    },
+    {
+      "epoch": 0.7783429890254816,
+      "grad_norm": 2.5791871547698975,
+      "learning_rate": 4.439263341340493e-05,
+      "loss": 1.9794,
+      "step": 33050
+    },
+    {
+      "epoch": 0.7785784937120249,
+      "grad_norm": 2.0614676475524902,
+      "learning_rate": 4.434553247609627e-05,
+      "loss": 1.8569,
+      "step": 33060
+    },
+    {
+      "epoch": 0.7788139983985681,
+      "grad_norm": 3.1220650672912598,
+      "learning_rate": 4.429843153878762e-05,
+      "loss": 2.0549,
+      "step": 33070
+    },
+    {
+      "epoch": 0.7790495030851113,
+      "grad_norm": NaN,
+      "learning_rate": 4.425133060147897e-05,
+      "loss": 1.9899,
+      "step": 33080
+    },
+    {
+      "epoch": 0.7792850077716547,
+      "grad_norm": 2.013601541519165,
+      "learning_rate": 4.4208939757901184e-05,
+      "loss": 1.9919,
+      "step": 33090
+    },
+    {
+      "epoch": 0.7795205124581979,
+      "grad_norm": 2.3044540882110596,
+      "learning_rate": 4.4161838820592534e-05,
+      "loss": 2.0695,
+      "step": 33100
+    },
+    {
+      "epoch": 0.7797560171447412,
+      "grad_norm": 2.1326963901519775,
+      "learning_rate": 4.411473788328388e-05,
+      "loss": 2.0311,
+      "step": 33110
+    },
+    {
+      "epoch": 0.7799915218312844,
+      "grad_norm": 2.112426280975342,
+      "learning_rate": 4.406763694597523e-05,
+      "loss": 1.8339,
+      "step": 33120
+    },
+    {
+      "epoch": 0.7802270265178277,
+      "grad_norm": 1.8452937602996826,
+      "learning_rate": 4.402053600866657e-05,
+      "loss": 2.0248,
+      "step": 33130
+    },
+    {
+      "epoch": 0.7804625312043709,
+      "grad_norm": 2.265781879425049,
+      "learning_rate": 4.397343507135792e-05,
+      "loss": 1.9383,
+      "step": 33140
+    },
+    {
+      "epoch": 0.7806980358909142,
+      "grad_norm": 2.641981601715088,
+      "learning_rate": 4.3926334134049265e-05,
+      "loss": 2.059,
+      "step": 33150
+    },
+    {
+      "epoch": 0.7809335405774575,
+      "grad_norm": 2.4268856048583984,
+      "learning_rate": 4.3879233196740616e-05,
+      "loss": 2.2207,
+      "step": 33160
+    },
+    {
+      "epoch": 0.7811690452640008,
+      "grad_norm": 2.2304317951202393,
+      "learning_rate": 4.3832132259431966e-05,
+      "loss": 2.1449,
+      "step": 33170
+    },
+    {
+      "epoch": 0.781404549950544,
+      "grad_norm": 2.062983274459839,
+      "learning_rate": 4.378503132212331e-05,
+      "loss": 1.8112,
+      "step": 33180
+    },
+    {
+      "epoch": 0.7816400546370873,
+      "grad_norm": 1.8598015308380127,
+      "learning_rate": 4.373793038481466e-05,
+      "loss": 1.8849,
+      "step": 33190
+    },
+    {
+      "epoch": 0.7818755593236305,
+      "grad_norm": 1.722225546836853,
+      "learning_rate": 4.369082944750601e-05,
+      "loss": 2.0335,
+      "step": 33200
+    },
+    {
+      "epoch": 0.7821110640101738,
+      "grad_norm": 4.0551958084106445,
+      "learning_rate": 4.364372851019736e-05,
+      "loss": 2.0743,
+      "step": 33210
+    },
+    {
+      "epoch": 0.782346568696717,
+      "grad_norm": 1.9605753421783447,
+      "learning_rate": 4.360133766661957e-05,
+      "loss": 2.0431,
+      "step": 33220
+    },
+    {
+      "epoch": 0.7825820733832604,
+      "grad_norm": 2.312108278274536,
+      "learning_rate": 4.3554236729310915e-05,
+      "loss": 1.9877,
+      "step": 33230
+    },
+    {
+      "epoch": 0.7828175780698036,
+      "grad_norm": 2.241905927658081,
+      "learning_rate": 4.3507135792002265e-05,
+      "loss": 2.0147,
+      "step": 33240
+    },
+    {
+      "epoch": 0.7830530827563469,
+      "grad_norm": 2.4801273345947266,
+      "learning_rate": 4.346003485469361e-05,
+      "loss": 1.9757,
+      "step": 33250
+    },
+    {
+      "epoch": 0.7832885874428901,
+      "grad_norm": 2.7878518104553223,
+      "learning_rate": 4.341293391738496e-05,
+      "loss": 2.0236,
+      "step": 33260
+    },
+    {
+      "epoch": 0.7835240921294334,
+      "grad_norm": 1.9572044610977173,
+      "learning_rate": 4.33658329800763e-05,
+      "loss": 2.1393,
+      "step": 33270
+    },
+    {
+      "epoch": 0.7837595968159766,
+      "grad_norm": 1.8226878643035889,
+      "learning_rate": 4.331873204276765e-05,
+      "loss": 1.9322,
+      "step": 33280
+    },
+    {
+      "epoch": 0.7839951015025199,
+      "grad_norm": 1.8301420211791992,
+      "learning_rate": 4.3271631105458996e-05,
+      "loss": 2.0656,
+      "step": 33290
+    },
+    {
+      "epoch": 0.7842306061890632,
+      "grad_norm": 3.304522752761841,
+      "learning_rate": 4.3224530168150346e-05,
+      "loss": 1.7974,
+      "step": 33300
+    },
+    {
+      "epoch": 0.7844661108756065,
+      "grad_norm": 2.201037645339966,
+      "learning_rate": 4.3177429230841696e-05,
+      "loss": 1.9841,
+      "step": 33310
+    },
+    {
+      "epoch": 0.7847016155621497,
+      "grad_norm": 1.9371416568756104,
+      "learning_rate": 4.3130328293533047e-05,
+      "loss": 2.1516,
+      "step": 33320
+    },
+    {
+      "epoch": 0.784937120248693,
+      "grad_norm": 2.4234302043914795,
+      "learning_rate": 4.308322735622439e-05,
+      "loss": 2.0129,
+      "step": 33330
+    },
+    {
+      "epoch": 0.7851726249352362,
+      "grad_norm": 1.8257781267166138,
+      "learning_rate": 4.303612641891574e-05,
+      "loss": 2.0591,
+      "step": 33340
+    },
+    {
+      "epoch": 0.7854081296217795,
+      "grad_norm": 2.209580659866333,
+      "learning_rate": 4.2989025481607084e-05,
+      "loss": 2.033,
+      "step": 33350
+    },
+    {
+      "epoch": 0.7856436343083227,
+      "grad_norm": 2.0317904949188232,
+      "learning_rate": 4.2941924544298434e-05,
+      "loss": 2.0874,
+      "step": 33360
+    },
+    {
+      "epoch": 0.785879138994866,
+      "grad_norm": 2.2737340927124023,
+      "learning_rate": 4.2894823606989784e-05,
+      "loss": 2.0869,
+      "step": 33370
+    },
+    {
+      "epoch": 0.7861146436814093,
+      "grad_norm": 1.857039213180542,
+      "learning_rate": 4.284772266968113e-05,
+      "loss": 2.1403,
+      "step": 33380
+    },
+    {
+      "epoch": 0.7863501483679525,
+      "grad_norm": 2.16298508644104,
+      "learning_rate": 4.280062173237248e-05,
+      "loss": 1.8499,
+      "step": 33390
+    },
+    {
+      "epoch": 0.7865856530544958,
+      "grad_norm": 2.2282299995422363,
+      "learning_rate": 4.275352079506382e-05,
+      "loss": 2.1735,
+      "step": 33400
+    },
+    {
+      "epoch": 0.786821157741039,
+      "grad_norm": 1.993444800376892,
+      "learning_rate": 4.270641985775517e-05,
+      "loss": 2.1354,
+      "step": 33410
+    },
+    {
+      "epoch": 0.7870566624275823,
+      "grad_norm": 2.0155577659606934,
+      "learning_rate": 4.2659318920446515e-05,
+      "loss": 1.7992,
+      "step": 33420
+    },
+    {
+      "epoch": 0.7872921671141255,
+      "grad_norm": 2.0637929439544678,
+      "learning_rate": 4.2612217983137865e-05,
+      "loss": 1.9629,
+      "step": 33430
+    },
+    {
+      "epoch": 0.7875276718006688,
+      "grad_norm": 2.51766300201416,
+      "learning_rate": 4.256511704582921e-05,
+      "loss": 1.8442,
+      "step": 33440
+    },
+    {
+      "epoch": 0.7877631764872121,
+      "grad_norm": 2.253438949584961,
+      "learning_rate": 4.2518016108520566e-05,
+      "loss": 1.9771,
+      "step": 33450
+    },
+    {
+      "epoch": 0.7879986811737554,
+      "grad_norm": 2.1923258304595947,
+      "learning_rate": 4.247091517121191e-05,
+      "loss": 2.2046,
+      "step": 33460
+    },
+    {
+      "epoch": 0.7882341858602986,
+      "grad_norm": 1.9330971240997314,
+      "learning_rate": 4.242381423390326e-05,
+      "loss": 1.9438,
+      "step": 33470
+    },
+    {
+      "epoch": 0.7884696905468419,
+      "grad_norm": 1.7735408544540405,
+      "learning_rate": 4.23767132965946e-05,
+      "loss": 2.0522,
+      "step": 33480
+    },
+    {
+      "epoch": 0.7887051952333851,
+      "grad_norm": 2.54807448387146,
+      "learning_rate": 4.232961235928595e-05,
+      "loss": 1.9661,
+      "step": 33490
+    },
+    {
+      "epoch": 0.7889406999199284,
+      "grad_norm": 1.8328182697296143,
+      "learning_rate": 4.22825114219773e-05,
+      "loss": 2.1688,
+      "step": 33500
+    },
+    {
+      "epoch": 0.7891762046064716,
+      "grad_norm": 2.4814724922180176,
+      "learning_rate": 4.223541048466865e-05,
+      "loss": 2.03,
+      "step": 33510
+    },
+    {
+      "epoch": 0.789411709293015,
+      "grad_norm": 2.5519399642944336,
+      "learning_rate": 4.218830954736e-05,
+      "loss": 2.1314,
+      "step": 33520
+    },
+    {
+      "epoch": 0.7896472139795582,
+      "grad_norm": 2.523111343383789,
+      "learning_rate": 4.214120861005134e-05,
+      "loss": 1.9807,
+      "step": 33530
+    },
+    {
+      "epoch": 0.7898827186661015,
+      "grad_norm": 2.3048107624053955,
+      "learning_rate": 4.209410767274269e-05,
+      "loss": 2.0443,
+      "step": 33540
+    },
+    {
+      "epoch": 0.7901182233526447,
+      "grad_norm": 2.162158966064453,
+      "learning_rate": 4.2047006735434034e-05,
+      "loss": 2.0237,
+      "step": 33550
+    },
+    {
+      "epoch": 0.790353728039188,
+      "grad_norm": 2.4160468578338623,
+      "learning_rate": 4.1999905798125385e-05,
+      "loss": 2.1968,
+      "step": 33560
+    },
+    {
+      "epoch": 0.7905892327257312,
+      "grad_norm": 1.7282012701034546,
+      "learning_rate": 4.195280486081673e-05,
+      "loss": 1.9573,
+      "step": 33570
+    },
+    {
+      "epoch": 0.7908247374122745,
+      "grad_norm": 2.083235025405884,
+      "learning_rate": 4.190570392350808e-05,
+      "loss": 2.0517,
+      "step": 33580
+    },
+    {
+      "epoch": 0.7910602420988178,
+      "grad_norm": 2.5902810096740723,
+      "learning_rate": 4.185860298619943e-05,
+      "loss": 2.072,
+      "step": 33590
+    },
+    {
+      "epoch": 0.7912957467853611,
+      "grad_norm": 2.1250321865081787,
+      "learning_rate": 4.181150204889078e-05,
+      "loss": 2.1724,
+      "step": 33600
+    },
+    {
+      "epoch": 0.7915312514719043,
+      "grad_norm": 2.0615923404693604,
+      "learning_rate": 4.176440111158212e-05,
+      "loss": 2.0236,
+      "step": 33610
+    },
+    {
+      "epoch": 0.7917667561584476,
+      "grad_norm": 2.1386961936950684,
+      "learning_rate": 4.171730017427347e-05,
+      "loss": 2.1057,
+      "step": 33620
+    },
+    {
+      "epoch": 0.7920022608449908,
+      "grad_norm": 2.5183324813842773,
+      "learning_rate": 4.1670199236964816e-05,
+      "loss": 2.0803,
+      "step": 33630
+    },
+    {
+      "epoch": 0.7922377655315341,
+      "grad_norm": 2.7656891345977783,
+      "learning_rate": 4.1623098299656166e-05,
+      "loss": 2.0875,
+      "step": 33640
+    },
+    {
+      "epoch": 0.7924732702180773,
+      "grad_norm": 1.7094897031784058,
+      "learning_rate": 4.157599736234751e-05,
+      "loss": 2.0225,
+      "step": 33650
+    },
+    {
+      "epoch": 0.7927087749046206,
+      "grad_norm": 2.0408098697662354,
+      "learning_rate": 4.152889642503886e-05,
+      "loss": 1.9889,
+      "step": 33660
+    },
+    {
+      "epoch": 0.7929442795911639,
+      "grad_norm": 2.6148180961608887,
+      "learning_rate": 4.148179548773021e-05,
+      "loss": 2.3067,
+      "step": 33670
+    },
+    {
+      "epoch": 0.7931797842777071,
+      "grad_norm": 1.6949024200439453,
+      "learning_rate": 4.1434694550421554e-05,
+      "loss": 2.2897,
+      "step": 33680
+    },
+    {
+      "epoch": 0.7934152889642504,
+      "grad_norm": 2.3153419494628906,
+      "learning_rate": 4.1387593613112904e-05,
+      "loss": 1.8178,
+      "step": 33690
+    },
+    {
+      "epoch": 0.7936507936507936,
+      "grad_norm": 2.5159363746643066,
+      "learning_rate": 4.134049267580425e-05,
+      "loss": 2.291,
+      "step": 33700
+    },
+    {
+      "epoch": 0.7938862983373369,
+      "grad_norm": 1.6296318769454956,
+      "learning_rate": 4.12933917384956e-05,
+      "loss": 1.9528,
+      "step": 33710
+    },
+    {
+      "epoch": 0.7941218030238801,
+      "grad_norm": 2.256610870361328,
+      "learning_rate": 4.124629080118694e-05,
+      "loss": 2.1584,
+      "step": 33720
+    },
+    {
+      "epoch": 0.7943573077104235,
+      "grad_norm": 1.8672707080841064,
+      "learning_rate": 4.11991898638783e-05,
+      "loss": 2.0653,
+      "step": 33730
+    },
+    {
+      "epoch": 0.7945928123969667,
+      "grad_norm": 3.621340751647949,
+      "learning_rate": 4.115208892656964e-05,
+      "loss": 2.2101,
+      "step": 33740
+    },
+    {
+      "epoch": 0.79482831708351,
+      "grad_norm": 2.5111987590789795,
+      "learning_rate": 4.110498798926099e-05,
+      "loss": 2.0825,
+      "step": 33750
+    },
+    {
+      "epoch": 0.7950638217700532,
+      "grad_norm": 2.316218614578247,
+      "learning_rate": 4.1057887051952335e-05,
+      "loss": 1.979,
+      "step": 33760
+    },
+    {
+      "epoch": 0.7952993264565965,
+      "grad_norm": 3.524585247039795,
+      "learning_rate": 4.1010786114643686e-05,
+      "loss": 1.9178,
+      "step": 33770
+    },
+    {
+      "epoch": 0.7955348311431397,
+      "grad_norm": 2.1990773677825928,
+      "learning_rate": 4.096368517733503e-05,
+      "loss": 2.0901,
+      "step": 33780
+    },
+    {
+      "epoch": 0.795770335829683,
+      "grad_norm": 3.369415760040283,
+      "learning_rate": 4.091658424002638e-05,
+      "loss": 1.7894,
+      "step": 33790
+    },
+    {
+      "epoch": 0.7960058405162262,
+      "grad_norm": 1.9027162790298462,
+      "learning_rate": 4.086948330271772e-05,
+      "loss": 2.0495,
+      "step": 33800
+    },
+    {
+      "epoch": 0.7962413452027696,
+      "grad_norm": 2.3677480220794678,
+      "learning_rate": 4.082238236540907e-05,
+      "loss": 2.124,
+      "step": 33810
+    },
+    {
+      "epoch": 0.7964768498893128,
+      "grad_norm": 2.0192174911499023,
+      "learning_rate": 4.077528142810042e-05,
+      "loss": 1.9715,
+      "step": 33820
+    },
+    {
+      "epoch": 0.7967123545758561,
+      "grad_norm": 2.7535741329193115,
+      "learning_rate": 4.072818049079177e-05,
+      "loss": 1.9513,
+      "step": 33830
+    },
+    {
+      "epoch": 0.7969478592623993,
+      "grad_norm": 1.9463825225830078,
+      "learning_rate": 4.068107955348312e-05,
+      "loss": 1.8706,
+      "step": 33840
+    },
+    {
+      "epoch": 0.7971833639489426,
+      "grad_norm": 2.1753220558166504,
+      "learning_rate": 4.063397861617446e-05,
+      "loss": 2.2774,
+      "step": 33850
+    },
+    {
+      "epoch": 0.7974188686354858,
+      "grad_norm": 1.5415219068527222,
+      "learning_rate": 4.058687767886581e-05,
+      "loss": 2.0952,
+      "step": 33860
+    },
+    {
+      "epoch": 0.7976543733220292,
+      "grad_norm": 2.0215823650360107,
+      "learning_rate": 4.053977674155716e-05,
+      "loss": 1.9228,
+      "step": 33870
+    },
+    {
+      "epoch": 0.7978898780085724,
+      "grad_norm": 1.6013972759246826,
+      "learning_rate": 4.049267580424851e-05,
+      "loss": 2.037,
+      "step": 33880
+    },
+    {
+      "epoch": 0.7981253826951157,
+      "grad_norm": 2.624075174331665,
+      "learning_rate": 4.0445574866939855e-05,
+      "loss": 1.9571,
+      "step": 33890
+    },
+    {
+      "epoch": 0.7983608873816589,
+      "grad_norm": 2.599543571472168,
+      "learning_rate": 4.0398473929631205e-05,
+      "loss": 2.0544,
+      "step": 33900
+    },
+    {
+      "epoch": 0.7985963920682022,
+      "grad_norm": 1.950036883354187,
+      "learning_rate": 4.035137299232255e-05,
+      "loss": 2.0248,
+      "step": 33910
+    },
+    {
+      "epoch": 0.7988318967547454,
+      "grad_norm": 2.2384231090545654,
+      "learning_rate": 4.03042720550139e-05,
+      "loss": 2.0692,
+      "step": 33920
+    },
+    {
+      "epoch": 0.7990674014412886,
+      "grad_norm": 1.6910631656646729,
+      "learning_rate": 4.025717111770524e-05,
+      "loss": 1.9093,
+      "step": 33930
+    },
+    {
+      "epoch": 0.799302906127832,
+      "grad_norm": 2.5212926864624023,
+      "learning_rate": 4.021007018039659e-05,
+      "loss": 1.9333,
+      "step": 33940
+    },
+    {
+      "epoch": 0.7995384108143752,
+      "grad_norm": 2.185635566711426,
+      "learning_rate": 4.0162969243087936e-05,
+      "loss": 2.2586,
+      "step": 33950
+    },
+    {
+      "epoch": 0.7997739155009185,
+      "grad_norm": 3.235745429992676,
+      "learning_rate": 4.0115868305779286e-05,
+      "loss": 2.0323,
+      "step": 33960
+    },
+    {
+      "epoch": 0.8000094201874617,
+      "grad_norm": 1.8148341178894043,
+      "learning_rate": 4.0068767368470636e-05,
+      "loss": 2.091,
+      "step": 33970
+    },
+    {
+      "epoch": 0.800244924874005,
+      "grad_norm": 2.3433852195739746,
+      "learning_rate": 4.002166643116198e-05,
+      "loss": 1.7543,
+      "step": 33980
+    },
+    {
+      "epoch": 0.8004804295605482,
+      "grad_norm": 1.957887053489685,
+      "learning_rate": 3.997456549385333e-05,
+      "loss": 1.9187,
+      "step": 33990
+    },
+    {
+      "epoch": 0.8007159342470915,
+      "grad_norm": 2.367794990539551,
+      "learning_rate": 3.9927464556544673e-05,
+      "loss": 2.0184,
+      "step": 34000
+    },
+    {
+      "epoch": 0.8009514389336347,
+      "grad_norm": 2.2443687915802,
+      "learning_rate": 3.9880363619236024e-05,
+      "loss": 1.9743,
+      "step": 34010
+    },
+    {
+      "epoch": 0.8011869436201781,
+      "grad_norm": 1.9915101528167725,
+      "learning_rate": 3.9833262681927374e-05,
+      "loss": 1.9113,
+      "step": 34020
+    },
+    {
+      "epoch": 0.8014224483067213,
+      "grad_norm": 2.523921489715576,
+      "learning_rate": 3.9786161744618724e-05,
+      "loss": 1.9004,
+      "step": 34030
+    },
+    {
+      "epoch": 0.8016579529932646,
+      "grad_norm": 2.2113091945648193,
+      "learning_rate": 3.973906080731007e-05,
+      "loss": 2.2037,
+      "step": 34040
+    },
+    {
+      "epoch": 0.8018934576798078,
+      "grad_norm": 1.9317169189453125,
+      "learning_rate": 3.969195987000142e-05,
+      "loss": 2.1463,
+      "step": 34050
+    },
+    {
+      "epoch": 0.8021289623663511,
+      "grad_norm": 1.804018259048462,
+      "learning_rate": 3.964485893269276e-05,
+      "loss": 2.1225,
+      "step": 34060
+    },
+    {
+      "epoch": 0.8023644670528943,
+      "grad_norm": 1.9248290061950684,
+      "learning_rate": 3.959775799538411e-05,
+      "loss": 2.0875,
+      "step": 34070
+    },
+    {
+      "epoch": 0.8025999717394376,
+      "grad_norm": 2.0950820446014404,
+      "learning_rate": 3.9550657058075455e-05,
+      "loss": 2.1594,
+      "step": 34080
+    },
+    {
+      "epoch": 0.8028354764259809,
+      "grad_norm": 2.13773775100708,
+      "learning_rate": 3.9503556120766805e-05,
+      "loss": 1.6516,
+      "step": 34090
+    },
+    {
+      "epoch": 0.8030709811125242,
+      "grad_norm": 1.8024755716323853,
+      "learning_rate": 3.945645518345815e-05,
+      "loss": 1.9598,
+      "step": 34100
+    },
+    {
+      "epoch": 0.8033064857990674,
+      "grad_norm": 2.8858230113983154,
+      "learning_rate": 3.94093542461495e-05,
+      "loss": 1.8878,
+      "step": 34110
+    },
+    {
+      "epoch": 0.8035419904856107,
+      "grad_norm": 2.173196315765381,
+      "learning_rate": 3.936225330884085e-05,
+      "loss": 1.8901,
+      "step": 34120
+    },
+    {
+      "epoch": 0.8037774951721539,
+      "grad_norm": 2.5591368675231934,
+      "learning_rate": 3.931515237153219e-05,
+      "loss": 1.82,
+      "step": 34130
+    },
+    {
+      "epoch": 0.8040129998586972,
+      "grad_norm": 2.587899923324585,
+      "learning_rate": 3.926805143422354e-05,
+      "loss": 2.0842,
+      "step": 34140
+    },
+    {
+      "epoch": 0.8042485045452404,
+      "grad_norm": 1.9660160541534424,
+      "learning_rate": 3.9220950496914886e-05,
+      "loss": 1.9926,
+      "step": 34150
+    },
+    {
+      "epoch": 0.8044840092317838,
+      "grad_norm": 2.291707754135132,
+      "learning_rate": 3.9173849559606243e-05,
+      "loss": 1.9393,
+      "step": 34160
+    },
+    {
+      "epoch": 0.804719513918327,
+      "grad_norm": 2.452038288116455,
+      "learning_rate": 3.912674862229759e-05,
+      "loss": 2.0658,
+      "step": 34170
+    },
+    {
+      "epoch": 0.8049550186048703,
+      "grad_norm": 1.9962801933288574,
+      "learning_rate": 3.907964768498894e-05,
+      "loss": 2.0453,
+      "step": 34180
+    },
+    {
+      "epoch": 0.8051905232914135,
+      "grad_norm": 2.2015650272369385,
+      "learning_rate": 3.903254674768028e-05,
+      "loss": 2.1626,
+      "step": 34190
+    },
+    {
+      "epoch": 0.8054260279779568,
+      "grad_norm": 2.1929237842559814,
+      "learning_rate": 3.898544581037163e-05,
+      "loss": 2.1904,
+      "step": 34200
+    },
+    {
+      "epoch": 0.8056615326645,
+      "grad_norm": 2.3259692192077637,
+      "learning_rate": 3.8938344873062974e-05,
+      "loss": 1.9968,
+      "step": 34210
+    },
+    {
+      "epoch": 0.8058970373510432,
+      "grad_norm": 2.7438056468963623,
+      "learning_rate": 3.8891243935754325e-05,
+      "loss": 2.2442,
+      "step": 34220
+    },
+    {
+      "epoch": 0.8061325420375866,
+      "grad_norm": 2.609450340270996,
+      "learning_rate": 3.884414299844567e-05,
+      "loss": 1.9798,
+      "step": 34230
+    },
+    {
+      "epoch": 0.8063680467241298,
+      "grad_norm": 2.7057836055755615,
+      "learning_rate": 3.879704206113702e-05,
+      "loss": 1.6244,
+      "step": 34240
+    },
+    {
+      "epoch": 0.8066035514106731,
+      "grad_norm": 2.6504015922546387,
+      "learning_rate": 3.874994112382836e-05,
+      "loss": 2.1608,
+      "step": 34250
+    },
+    {
+      "epoch": 0.8068390560972163,
+      "grad_norm": 2.614208936691284,
+      "learning_rate": 3.870284018651971e-05,
+      "loss": 2.085,
+      "step": 34260
+    },
+    {
+      "epoch": 0.8070745607837596,
+      "grad_norm": 2.1055519580841064,
+      "learning_rate": 3.865573924921106e-05,
+      "loss": 2.0212,
+      "step": 34270
+    },
+    {
+      "epoch": 0.8073100654703028,
+      "grad_norm": 2.314775228500366,
+      "learning_rate": 3.8608638311902406e-05,
+      "loss": 2.0582,
+      "step": 34280
+    },
+    {
+      "epoch": 0.8075455701568461,
+      "grad_norm": 2.320479393005371,
+      "learning_rate": 3.8561537374593756e-05,
+      "loss": 1.8997,
+      "step": 34290
+    },
+    {
+      "epoch": 0.8077810748433893,
+      "grad_norm": 2.615222692489624,
+      "learning_rate": 3.8514436437285106e-05,
+      "loss": 2.1215,
+      "step": 34300
+    },
+    {
+      "epoch": 0.8080165795299327,
+      "grad_norm": 1.9585306644439697,
+      "learning_rate": 3.8467335499976456e-05,
+      "loss": 1.8919,
+      "step": 34310
+    },
+    {
+      "epoch": 0.8082520842164759,
+      "grad_norm": 1.8004975318908691,
+      "learning_rate": 3.84202345626678e-05,
+      "loss": 2.1203,
+      "step": 34320
+    },
+    {
+      "epoch": 0.8084875889030192,
+      "grad_norm": 1.973084568977356,
+      "learning_rate": 3.837313362535915e-05,
+      "loss": 2.0844,
+      "step": 34330
+    },
+    {
+      "epoch": 0.8087230935895624,
+      "grad_norm": 2.2893195152282715,
+      "learning_rate": 3.8326032688050494e-05,
+      "loss": 1.9385,
+      "step": 34340
+    },
+    {
+      "epoch": 0.8089585982761057,
+      "grad_norm": 2.5790085792541504,
+      "learning_rate": 3.8278931750741844e-05,
+      "loss": 1.9747,
+      "step": 34350
+    },
+    {
+      "epoch": 0.8091941029626489,
+      "grad_norm": 2.1595449447631836,
+      "learning_rate": 3.823183081343319e-05,
+      "loss": 2.1628,
+      "step": 34360
+    },
+    {
+      "epoch": 0.8094296076491923,
+      "grad_norm": 2.5433499813079834,
+      "learning_rate": 3.818472987612454e-05,
+      "loss": 1.9291,
+      "step": 34370
+    },
+    {
+      "epoch": 0.8096651123357355,
+      "grad_norm": 2.732644557952881,
+      "learning_rate": 3.813762893881588e-05,
+      "loss": 2.1055,
+      "step": 34380
+    },
+    {
+      "epoch": 0.8099006170222788,
+      "grad_norm": 1.8708646297454834,
+      "learning_rate": 3.809052800150723e-05,
+      "loss": 1.9993,
+      "step": 34390
+    },
+    {
+      "epoch": 0.810136121708822,
+      "grad_norm": 1.994516134262085,
+      "learning_rate": 3.8043427064198575e-05,
+      "loss": 2.0301,
+      "step": 34400
+    },
+    {
+      "epoch": 0.8103716263953653,
+      "grad_norm": 2.354677677154541,
+      "learning_rate": 3.7996326126889925e-05,
+      "loss": 2.0536,
+      "step": 34410
+    },
+    {
+      "epoch": 0.8106071310819085,
+      "grad_norm": 1.9902782440185547,
+      "learning_rate": 3.7949225189581275e-05,
+      "loss": 2.0342,
+      "step": 34420
+    },
+    {
+      "epoch": 0.8108426357684518,
+      "grad_norm": 2.1519672870635986,
+      "learning_rate": 3.790212425227262e-05,
+      "loss": 2.0471,
+      "step": 34430
+    },
+    {
+      "epoch": 0.811078140454995,
+      "grad_norm": 2.525325059890747,
+      "learning_rate": 3.785502331496397e-05,
+      "loss": 1.9082,
+      "step": 34440
+    },
+    {
+      "epoch": 0.8113136451415384,
+      "grad_norm": 1.9717092514038086,
+      "learning_rate": 3.780792237765532e-05,
+      "loss": 2.0308,
+      "step": 34450
+    },
+    {
+      "epoch": 0.8115491498280816,
+      "grad_norm": 2.371563196182251,
+      "learning_rate": 3.776082144034667e-05,
+      "loss": 2.2107,
+      "step": 34460
+    },
+    {
+      "epoch": 0.8117846545146249,
+      "grad_norm": 2.0133941173553467,
+      "learning_rate": 3.771372050303801e-05,
+      "loss": 2.2433,
+      "step": 34470
+    },
+    {
+      "epoch": 0.8120201592011681,
+      "grad_norm": 2.075589418411255,
+      "learning_rate": 3.766661956572936e-05,
+      "loss": 2.0989,
+      "step": 34480
+    },
+    {
+      "epoch": 0.8122556638877113,
+      "grad_norm": 2.919072389602661,
+      "learning_rate": 3.761951862842071e-05,
+      "loss": 2.1238,
+      "step": 34490
+    },
+    {
+      "epoch": 0.8124911685742546,
+      "grad_norm": 2.5039098262786865,
+      "learning_rate": 3.757241769111206e-05,
+      "loss": 2.1452,
+      "step": 34500
+    },
+    {
+      "epoch": 0.8127266732607978,
+      "grad_norm": 2.3447000980377197,
+      "learning_rate": 3.75253167538034e-05,
+      "loss": 2.1173,
+      "step": 34510
+    },
+    {
+      "epoch": 0.8129621779473412,
+      "grad_norm": 2.079946279525757,
+      "learning_rate": 3.747821581649475e-05,
+      "loss": 1.7781,
+      "step": 34520
+    },
+    {
+      "epoch": 0.8131976826338844,
+      "grad_norm": 2.0972554683685303,
+      "learning_rate": 3.7431114879186094e-05,
+      "loss": 1.8383,
+      "step": 34530
+    },
+    {
+      "epoch": 0.8134331873204277,
+      "grad_norm": 1.8633981943130493,
+      "learning_rate": 3.7384013941877444e-05,
+      "loss": 1.967,
+      "step": 34540
+    },
+    {
+      "epoch": 0.8136686920069709,
+      "grad_norm": 1.938206434249878,
+      "learning_rate": 3.733691300456879e-05,
+      "loss": 2.0227,
+      "step": 34550
+    },
+    {
+      "epoch": 0.8139041966935142,
+      "grad_norm": 2.096994400024414,
+      "learning_rate": 3.728981206726014e-05,
+      "loss": 2.109,
+      "step": 34560
+    },
+    {
+      "epoch": 0.8141397013800574,
+      "grad_norm": 2.675692319869995,
+      "learning_rate": 3.724271112995149e-05,
+      "loss": 2.0283,
+      "step": 34570
+    },
+    {
+      "epoch": 0.8143752060666007,
+      "grad_norm": 2.3348910808563232,
+      "learning_rate": 3.719561019264284e-05,
+      "loss": 2.0057,
+      "step": 34580
+    },
+    {
+      "epoch": 0.814610710753144,
+      "grad_norm": 1.5937950611114502,
+      "learning_rate": 3.714850925533418e-05,
+      "loss": 1.9629,
+      "step": 34590
+    },
+    {
+      "epoch": 0.8148462154396873,
+      "grad_norm": 2.59680438041687,
+      "learning_rate": 3.710140831802553e-05,
+      "loss": 1.8433,
+      "step": 34600
+    },
+    {
+      "epoch": 0.8150817201262305,
+      "grad_norm": 2.0707480907440186,
+      "learning_rate": 3.705430738071688e-05,
+      "loss": 1.9767,
+      "step": 34610
+    },
+    {
+      "epoch": 0.8153172248127738,
+      "grad_norm": 2.1386687755584717,
+      "learning_rate": 3.7007206443408226e-05,
+      "loss": 1.9353,
+      "step": 34620
+    },
+    {
+      "epoch": 0.815552729499317,
+      "grad_norm": 2.207897663116455,
+      "learning_rate": 3.6960105506099576e-05,
+      "loss": 2.0739,
+      "step": 34630
+    },
+    {
+      "epoch": 0.8157882341858603,
+      "grad_norm": 1.8906501531600952,
+      "learning_rate": 3.691300456879092e-05,
+      "loss": 2.0127,
+      "step": 34640
+    },
+    {
+      "epoch": 0.8160237388724035,
+      "grad_norm": 2.1605560779571533,
+      "learning_rate": 3.686590363148227e-05,
+      "loss": 1.8724,
+      "step": 34650
+    },
+    {
+      "epoch": 0.8162592435589469,
+      "grad_norm": 2.182072401046753,
+      "learning_rate": 3.681880269417361e-05,
+      "loss": 1.968,
+      "step": 34660
+    },
+    {
+      "epoch": 0.8164947482454901,
+      "grad_norm": 2.064140796661377,
+      "learning_rate": 3.6771701756864964e-05,
+      "loss": 2.0005,
+      "step": 34670
+    },
+    {
+      "epoch": 0.8167302529320334,
+      "grad_norm": 2.037813186645508,
+      "learning_rate": 3.672460081955631e-05,
+      "loss": 2.0115,
+      "step": 34680
+    },
+    {
+      "epoch": 0.8169657576185766,
+      "grad_norm": 2.205673933029175,
+      "learning_rate": 3.667749988224766e-05,
+      "loss": 1.8754,
+      "step": 34690
+    },
+    {
+      "epoch": 0.8172012623051199,
+      "grad_norm": 2.1787984371185303,
+      "learning_rate": 3.6630398944939e-05,
+      "loss": 2.0636,
+      "step": 34700
+    },
+    {
+      "epoch": 0.8174367669916631,
+      "grad_norm": 2.24696946144104,
+      "learning_rate": 3.658329800763035e-05,
+      "loss": 2.071,
+      "step": 34710
+    },
+    {
+      "epoch": 0.8176722716782064,
+      "grad_norm": 1.690644383430481,
+      "learning_rate": 3.65361970703217e-05,
+      "loss": 1.9747,
+      "step": 34720
+    },
+    {
+      "epoch": 0.8179077763647497,
+      "grad_norm": 2.699493646621704,
+      "learning_rate": 3.648909613301305e-05,
+      "loss": 2.1167,
+      "step": 34730
+    },
+    {
+      "epoch": 0.818143281051293,
+      "grad_norm": 2.260894536972046,
+      "learning_rate": 3.6441995195704395e-05,
+      "loss": 1.8883,
+      "step": 34740
+    },
+    {
+      "epoch": 0.8183787857378362,
+      "grad_norm": 1.964855670928955,
+      "learning_rate": 3.6394894258395745e-05,
+      "loss": 2.0409,
+      "step": 34750
+    },
+    {
+      "epoch": 0.8186142904243795,
+      "grad_norm": 2.360180616378784,
+      "learning_rate": 3.6347793321087095e-05,
+      "loss": 1.9799,
+      "step": 34760
+    },
+    {
+      "epoch": 0.8188497951109227,
+      "grad_norm": 2.9807870388031006,
+      "learning_rate": 3.630069238377844e-05,
+      "loss": 2.2196,
+      "step": 34770
+    },
+    {
+      "epoch": 0.8190852997974659,
+      "grad_norm": 2.496344566345215,
+      "learning_rate": 3.625359144646979e-05,
+      "loss": 1.8034,
+      "step": 34780
+    },
+    {
+      "epoch": 0.8193208044840092,
+      "grad_norm": 2.7705297470092773,
+      "learning_rate": 3.620649050916113e-05,
+      "loss": 1.8999,
+      "step": 34790
+    },
+    {
+      "epoch": 0.8195563091705524,
+      "grad_norm": 2.4351303577423096,
+      "learning_rate": 3.615938957185248e-05,
+      "loss": 1.8937,
+      "step": 34800
+    },
+    {
+      "epoch": 0.8197918138570958,
+      "grad_norm": 2.3407814502716064,
+      "learning_rate": 3.6112288634543826e-05,
+      "loss": 2.2182,
+      "step": 34810
+    },
+    {
+      "epoch": 0.820027318543639,
+      "grad_norm": 1.9275591373443604,
+      "learning_rate": 3.606518769723518e-05,
+      "loss": 2.0223,
+      "step": 34820
+    },
+    {
+      "epoch": 0.8202628232301823,
+      "grad_norm": 2.0065784454345703,
+      "learning_rate": 3.601808675992652e-05,
+      "loss": 2.2715,
+      "step": 34830
+    },
+    {
+      "epoch": 0.8204983279167255,
+      "grad_norm": 2.351534366607666,
+      "learning_rate": 3.597098582261787e-05,
+      "loss": 2.1069,
+      "step": 34840
+    },
+    {
+      "epoch": 0.8207338326032688,
+      "grad_norm": 2.254944086074829,
+      "learning_rate": 3.5923884885309214e-05,
+      "loss": 2.1348,
+      "step": 34850
+    },
+    {
+      "epoch": 0.820969337289812,
+      "grad_norm": 4.640080451965332,
+      "learning_rate": 3.587678394800057e-05,
+      "loss": 1.981,
+      "step": 34860
+    },
+    {
+      "epoch": 0.8212048419763553,
+      "grad_norm": 1.7675470113754272,
+      "learning_rate": 3.5829683010691914e-05,
+      "loss": 1.9438,
+      "step": 34870
+    },
+    {
+      "epoch": 0.8214403466628986,
+      "grad_norm": 1.8417401313781738,
+      "learning_rate": 3.5782582073383265e-05,
+      "loss": 1.9817,
+      "step": 34880
+    },
+    {
+      "epoch": 0.8216758513494419,
+      "grad_norm": 2.11193585395813,
+      "learning_rate": 3.573548113607461e-05,
+      "loss": 2.0047,
+      "step": 34890
+    },
+    {
+      "epoch": 0.8219113560359851,
+      "grad_norm": 1.9333069324493408,
+      "learning_rate": 3.568838019876596e-05,
+      "loss": 1.9611,
+      "step": 34900
+    },
+    {
+      "epoch": 0.8221468607225284,
+      "grad_norm": 1.8805556297302246,
+      "learning_rate": 3.564127926145731e-05,
+      "loss": 1.9072,
+      "step": 34910
+    },
+    {
+      "epoch": 0.8223823654090716,
+      "grad_norm": 2.1735150814056396,
+      "learning_rate": 3.559417832414865e-05,
+      "loss": 1.8782,
+      "step": 34920
+    },
+    {
+      "epoch": 0.8226178700956149,
+      "grad_norm": 2.4724745750427246,
+      "learning_rate": 3.554707738684e-05,
+      "loss": 2.0225,
+      "step": 34930
+    },
+    {
+      "epoch": 0.8228533747821581,
+      "grad_norm": 2.609894037246704,
+      "learning_rate": 3.5499976449531346e-05,
+      "loss": 2.1728,
+      "step": 34940
+    },
+    {
+      "epoch": 0.8230888794687015,
+      "grad_norm": 2.194990634918213,
+      "learning_rate": 3.5452875512222696e-05,
+      "loss": 1.8967,
+      "step": 34950
+    },
+    {
+      "epoch": 0.8233243841552447,
+      "grad_norm": 1.8386231660842896,
+      "learning_rate": 3.540577457491404e-05,
+      "loss": 2.0285,
+      "step": 34960
+    },
+    {
+      "epoch": 0.823559888841788,
+      "grad_norm": 2.252108573913574,
+      "learning_rate": 3.535867363760539e-05,
+      "loss": 2.0548,
+      "step": 34970
+    },
+    {
+      "epoch": 0.8237953935283312,
+      "grad_norm": 1.995373249053955,
+      "learning_rate": 3.531157270029673e-05,
+      "loss": 2.0336,
+      "step": 34980
+    },
+    {
+      "epoch": 0.8240308982148745,
+      "grad_norm": 2.212472438812256,
+      "learning_rate": 3.526447176298808e-05,
+      "loss": 2.0237,
+      "step": 34990
+    },
+    {
+      "epoch": 0.8242664029014177,
+      "grad_norm": 2.2565536499023438,
+      "learning_rate": 3.5217370825679434e-05,
+      "loss": 2.2211,
+      "step": 35000
+    },
+    {
+      "epoch": 0.824501907587961,
+      "grad_norm": 2.493603229522705,
+      "learning_rate": 3.5170269888370784e-05,
+      "loss": 2.1334,
+      "step": 35010
+    },
+    {
+      "epoch": 0.8247374122745043,
+      "grad_norm": 2.159635305404663,
+      "learning_rate": 3.512316895106213e-05,
+      "loss": 1.9526,
+      "step": 35020
+    },
+    {
+      "epoch": 0.8249729169610476,
+      "grad_norm": 1.947871208190918,
+      "learning_rate": 3.507606801375348e-05,
+      "loss": 1.9906,
+      "step": 35030
+    },
+    {
+      "epoch": 0.8252084216475908,
+      "grad_norm": 2.8672406673431396,
+      "learning_rate": 3.502896707644482e-05,
+      "loss": 2.2269,
+      "step": 35040
+    },
+    {
+      "epoch": 0.8254439263341341,
+      "grad_norm": 2.212104082107544,
+      "learning_rate": 3.498186613913617e-05,
+      "loss": 1.9107,
+      "step": 35050
+    },
+    {
+      "epoch": 0.8256794310206773,
+      "grad_norm": 2.2118680477142334,
+      "learning_rate": 3.493476520182752e-05,
+      "loss": 2.1372,
+      "step": 35060
+    },
+    {
+      "epoch": 0.8259149357072205,
+      "grad_norm": 2.289013385772705,
+      "learning_rate": 3.4887664264518865e-05,
+      "loss": 1.9752,
+      "step": 35070
+    },
+    {
+      "epoch": 0.8261504403937638,
+      "grad_norm": 2.4700567722320557,
+      "learning_rate": 3.4840563327210215e-05,
+      "loss": 2.0911,
+      "step": 35080
+    },
+    {
+      "epoch": 0.826385945080307,
+      "grad_norm": 1.9941130876541138,
+      "learning_rate": 3.479346238990156e-05,
+      "loss": 1.8645,
+      "step": 35090
+    },
+    {
+      "epoch": 0.8266214497668504,
+      "grad_norm": 2.4892849922180176,
+      "learning_rate": 3.474636145259291e-05,
+      "loss": 1.8108,
+      "step": 35100
+    },
+    {
+      "epoch": 0.8268569544533936,
+      "grad_norm": 1.7279706001281738,
+      "learning_rate": 3.469926051528425e-05,
+      "loss": 1.8917,
+      "step": 35110
+    },
+    {
+      "epoch": 0.8270924591399369,
+      "grad_norm": 1.975010871887207,
+      "learning_rate": 3.46521595779756e-05,
+      "loss": 1.9208,
+      "step": 35120
+    },
+    {
+      "epoch": 0.8273279638264801,
+      "grad_norm": 2.1341896057128906,
+      "learning_rate": 3.4605058640666946e-05,
+      "loss": 2.0969,
+      "step": 35130
+    },
+    {
+      "epoch": 0.8275634685130234,
+      "grad_norm": 2.393791437149048,
+      "learning_rate": 3.4557957703358296e-05,
+      "loss": 1.9602,
+      "step": 35140
+    },
+    {
+      "epoch": 0.8277989731995666,
+      "grad_norm": 2.1387171745300293,
+      "learning_rate": 3.4510856766049647e-05,
+      "loss": 2.0583,
+      "step": 35150
+    },
+    {
+      "epoch": 0.82803447788611,
+      "grad_norm": 1.9722431898117065,
+      "learning_rate": 3.4463755828741e-05,
+      "loss": 2.0649,
+      "step": 35160
+    },
+    {
+      "epoch": 0.8282699825726532,
+      "grad_norm": 2.32572340965271,
+      "learning_rate": 3.441665489143234e-05,
+      "loss": 2.1126,
+      "step": 35170
+    },
+    {
+      "epoch": 0.8285054872591965,
+      "grad_norm": 2.5280470848083496,
+      "learning_rate": 3.436955395412369e-05,
+      "loss": 2.0072,
+      "step": 35180
+    },
+    {
+      "epoch": 0.8287409919457397,
+      "grad_norm": 2.0874745845794678,
+      "learning_rate": 3.4322453016815034e-05,
+      "loss": 2.0661,
+      "step": 35190
+    },
+    {
+      "epoch": 0.828976496632283,
+      "grad_norm": 3.3035924434661865,
+      "learning_rate": 3.4275352079506384e-05,
+      "loss": 1.9418,
+      "step": 35200
+    },
+    {
+      "epoch": 0.8292120013188262,
+      "grad_norm": 2.1365528106689453,
+      "learning_rate": 3.4228251142197735e-05,
+      "loss": 1.8834,
+      "step": 35210
+    },
+    {
+      "epoch": 0.8294475060053695,
+      "grad_norm": 1.7832911014556885,
+      "learning_rate": 3.418115020488908e-05,
+      "loss": 1.9537,
+      "step": 35220
+    },
+    {
+      "epoch": 0.8296830106919127,
+      "grad_norm": 2.0132415294647217,
+      "learning_rate": 3.413404926758043e-05,
+      "loss": 2.0083,
+      "step": 35230
+    },
+    {
+      "epoch": 0.8299185153784561,
+      "grad_norm": 2.6595003604888916,
+      "learning_rate": 3.408694833027177e-05,
+      "loss": 2.0959,
+      "step": 35240
+    },
+    {
+      "epoch": 0.8301540200649993,
+      "grad_norm": 1.921191692352295,
+      "learning_rate": 3.403984739296312e-05,
+      "loss": 1.9271,
+      "step": 35250
+    },
+    {
+      "epoch": 0.8303895247515426,
+      "grad_norm": 2.7424440383911133,
+      "learning_rate": 3.3992746455654465e-05,
+      "loss": 2.0904,
+      "step": 35260
+    },
+    {
+      "epoch": 0.8306250294380858,
+      "grad_norm": 5.584036827087402,
+      "learning_rate": 3.3945645518345816e-05,
+      "loss": 2.0697,
+      "step": 35270
+    },
+    {
+      "epoch": 0.8308605341246291,
+      "grad_norm": 2.842258930206299,
+      "learning_rate": 3.389854458103716e-05,
+      "loss": 1.9869,
+      "step": 35280
+    },
+    {
+      "epoch": 0.8310960388111723,
+      "grad_norm": 1.9061213731765747,
+      "learning_rate": 3.3851443643728516e-05,
+      "loss": 2.3207,
+      "step": 35290
+    },
+    {
+      "epoch": 0.8313315434977157,
+      "grad_norm": 2.775942087173462,
+      "learning_rate": 3.380434270641986e-05,
+      "loss": 1.9147,
+      "step": 35300
+    },
+    {
+      "epoch": 0.8315670481842589,
+      "grad_norm": 3.2270853519439697,
+      "learning_rate": 3.375724176911121e-05,
+      "loss": 2.175,
+      "step": 35310
+    },
+    {
+      "epoch": 0.8318025528708022,
+      "grad_norm": 1.8016290664672852,
+      "learning_rate": 3.371014083180255e-05,
+      "loss": 1.8835,
+      "step": 35320
+    },
+    {
+      "epoch": 0.8320380575573454,
+      "grad_norm": 2.776650905609131,
+      "learning_rate": 3.3663039894493904e-05,
+      "loss": 1.7118,
+      "step": 35330
+    },
+    {
+      "epoch": 0.8322735622438886,
+      "grad_norm": 2.003655433654785,
+      "learning_rate": 3.361593895718525e-05,
+      "loss": 1.9875,
+      "step": 35340
+    },
+    {
+      "epoch": 0.8325090669304319,
+      "grad_norm": 2.3793911933898926,
+      "learning_rate": 3.35688380198766e-05,
+      "loss": 2.1075,
+      "step": 35350
+    },
+    {
+      "epoch": 0.8327445716169751,
+      "grad_norm": 2.2091314792633057,
+      "learning_rate": 3.352173708256795e-05,
+      "loss": 2.068,
+      "step": 35360
+    },
+    {
+      "epoch": 0.8329800763035184,
+      "grad_norm": 2.356450319290161,
+      "learning_rate": 3.347463614525929e-05,
+      "loss": 2.0795,
+      "step": 35370
+    },
+    {
+      "epoch": 0.8332155809900617,
+      "grad_norm": 1.9805959463119507,
+      "learning_rate": 3.342753520795064e-05,
+      "loss": 2.0017,
+      "step": 35380
+    },
+    {
+      "epoch": 0.833451085676605,
+      "grad_norm": 1.8965378999710083,
+      "learning_rate": 3.3380434270641985e-05,
+      "loss": 2.0095,
+      "step": 35390
+    },
+    {
+      "epoch": 0.8336865903631482,
+      "grad_norm": 2.3308956623077393,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 1.6953,
+      "step": 35400
+    },
+    {
+      "epoch": 0.8339220950496915,
+      "grad_norm": 1.7506887912750244,
+      "learning_rate": 3.328623239602468e-05,
+      "loss": 2.011,
+      "step": 35410
+    },
+    {
+      "epoch": 0.8341575997362347,
+      "grad_norm": 1.6992536783218384,
+      "learning_rate": 3.323913145871603e-05,
+      "loss": 1.9904,
+      "step": 35420
+    },
+    {
+      "epoch": 0.834393104422778,
+      "grad_norm": 2.503995418548584,
+      "learning_rate": 3.319203052140738e-05,
+      "loss": 2.0786,
+      "step": 35430
+    },
+    {
+      "epoch": 0.8346286091093212,
+      "grad_norm": 1.9842063188552856,
+      "learning_rate": 3.314492958409873e-05,
+      "loss": 1.9527,
+      "step": 35440
+    },
+    {
+      "epoch": 0.8348641137958646,
+      "grad_norm": 1.7640032768249512,
+      "learning_rate": 3.309782864679007e-05,
+      "loss": 1.8065,
+      "step": 35450
+    },
+    {
+      "epoch": 0.8350996184824078,
+      "grad_norm": 2.2963879108428955,
+      "learning_rate": 3.305072770948142e-05,
+      "loss": 1.8249,
+      "step": 35460
+    },
+    {
+      "epoch": 0.8353351231689511,
+      "grad_norm": 1.9893380403518677,
+      "learning_rate": 3.3003626772172766e-05,
+      "loss": 2.1082,
+      "step": 35470
+    },
+    {
+      "epoch": 0.8355706278554943,
+      "grad_norm": 1.7458503246307373,
+      "learning_rate": 3.2956525834864117e-05,
+      "loss": 1.9985,
+      "step": 35480
+    },
+    {
+      "epoch": 0.8358061325420376,
+      "grad_norm": 1.997781753540039,
+      "learning_rate": 3.290942489755546e-05,
+      "loss": 2.1124,
+      "step": 35490
+    },
+    {
+      "epoch": 0.8360416372285808,
+      "grad_norm": 3.002406358718872,
+      "learning_rate": 3.286232396024681e-05,
+      "loss": 2.0895,
+      "step": 35500
+    },
+    {
+      "epoch": 0.8362771419151241,
+      "grad_norm": 1.5130270719528198,
+      "learning_rate": 3.281522302293816e-05,
+      "loss": 1.8941,
+      "step": 35510
+    },
+    {
+      "epoch": 0.8365126466016674,
+      "grad_norm": 1.7578104734420776,
+      "learning_rate": 3.2768122085629504e-05,
+      "loss": 2.0347,
+      "step": 35520
+    },
+    {
+      "epoch": 0.8367481512882107,
+      "grad_norm": 2.7366182804107666,
+      "learning_rate": 3.2721021148320854e-05,
+      "loss": 1.8944,
+      "step": 35530
+    },
+    {
+      "epoch": 0.8369836559747539,
+      "grad_norm": 2.484748601913452,
+      "learning_rate": 3.26739202110122e-05,
+      "loss": 1.9773,
+      "step": 35540
+    },
+    {
+      "epoch": 0.8372191606612972,
+      "grad_norm": 2.0619559288024902,
+      "learning_rate": 3.262681927370355e-05,
+      "loss": 2.0534,
+      "step": 35550
+    },
+    {
+      "epoch": 0.8374546653478404,
+      "grad_norm": 2.169790744781494,
+      "learning_rate": 3.257971833639489e-05,
+      "loss": 1.9748,
+      "step": 35560
+    },
+    {
+      "epoch": 0.8376901700343837,
+      "grad_norm": 2.2871201038360596,
+      "learning_rate": 3.253261739908625e-05,
+      "loss": 2.1623,
+      "step": 35570
+    },
+    {
+      "epoch": 0.8379256747209269,
+      "grad_norm": 2.192305326461792,
+      "learning_rate": 3.248551646177759e-05,
+      "loss": 1.9842,
+      "step": 35580
+    },
+    {
+      "epoch": 0.8381611794074703,
+      "grad_norm": 1.6627674102783203,
+      "learning_rate": 3.243841552446894e-05,
+      "loss": 2.0446,
+      "step": 35590
+    },
+    {
+      "epoch": 0.8383966840940135,
+      "grad_norm": 1.8032068014144897,
+      "learning_rate": 3.2391314587160286e-05,
+      "loss": 1.9092,
+      "step": 35600
+    },
+    {
+      "epoch": 0.8386321887805568,
+      "grad_norm": 2.1932010650634766,
+      "learning_rate": 3.2344213649851636e-05,
+      "loss": 2.0882,
+      "step": 35610
+    },
+    {
+      "epoch": 0.8388676934671,
+      "grad_norm": 2.2689177989959717,
+      "learning_rate": 3.229711271254298e-05,
+      "loss": 2.0097,
+      "step": 35620
+    },
+    {
+      "epoch": 0.8391031981536432,
+      "grad_norm": 1.843926191329956,
+      "learning_rate": 3.225001177523433e-05,
+      "loss": 2.1152,
+      "step": 35630
+    },
+    {
+      "epoch": 0.8393387028401865,
+      "grad_norm": 1.8309720754623413,
+      "learning_rate": 3.220291083792567e-05,
+      "loss": 1.7824,
+      "step": 35640
+    },
+    {
+      "epoch": 0.8395742075267297,
+      "grad_norm": 1.8428617715835571,
+      "learning_rate": 3.215580990061702e-05,
+      "loss": 2.0781,
+      "step": 35650
+    },
+    {
+      "epoch": 0.839809712213273,
+      "grad_norm": 3.1680748462677,
+      "learning_rate": 3.2108708963308374e-05,
+      "loss": 2.0001,
+      "step": 35660
+    },
+    {
+      "epoch": 0.8400452168998163,
+      "grad_norm": 2.44405198097229,
+      "learning_rate": 3.206160802599972e-05,
+      "loss": 1.9798,
+      "step": 35670
+    },
+    {
+      "epoch": 0.8402807215863596,
+      "grad_norm": 2.168524980545044,
+      "learning_rate": 3.201450708869107e-05,
+      "loss": 1.7701,
+      "step": 35680
+    },
+    {
+      "epoch": 0.8405162262729028,
+      "grad_norm": 2.2438528537750244,
+      "learning_rate": 3.196740615138241e-05,
+      "loss": 2.1856,
+      "step": 35690
+    },
+    {
+      "epoch": 0.8407517309594461,
+      "grad_norm": 3.037527084350586,
+      "learning_rate": 3.192030521407376e-05,
+      "loss": 2.1172,
+      "step": 35700
+    },
+    {
+      "epoch": 0.8409872356459893,
+      "grad_norm": 2.3657798767089844,
+      "learning_rate": 3.187320427676511e-05,
+      "loss": 2.2592,
+      "step": 35710
+    },
+    {
+      "epoch": 0.8412227403325326,
+      "grad_norm": 2.6857681274414062,
+      "learning_rate": 3.182610333945646e-05,
+      "loss": 1.8174,
+      "step": 35720
+    },
+    {
+      "epoch": 0.8414582450190758,
+      "grad_norm": 2.249049186706543,
+      "learning_rate": 3.1779002402147805e-05,
+      "loss": 2.0437,
+      "step": 35730
+    },
+    {
+      "epoch": 0.8416937497056192,
+      "grad_norm": 2.1826348304748535,
+      "learning_rate": 3.1731901464839155e-05,
+      "loss": 2.1848,
+      "step": 35740
+    },
+    {
+      "epoch": 0.8419292543921624,
+      "grad_norm": 2.04245924949646,
+      "learning_rate": 3.16848005275305e-05,
+      "loss": 1.8733,
+      "step": 35750
+    },
+    {
+      "epoch": 0.8421647590787057,
+      "grad_norm": 2.0764975547790527,
+      "learning_rate": 3.163769959022185e-05,
+      "loss": 1.9855,
+      "step": 35760
+    },
+    {
+      "epoch": 0.8424002637652489,
+      "grad_norm": 2.581373453140259,
+      "learning_rate": 3.159059865291319e-05,
+      "loss": 1.7349,
+      "step": 35770
+    },
+    {
+      "epoch": 0.8426357684517922,
+      "grad_norm": 2.6888554096221924,
+      "learning_rate": 3.154349771560454e-05,
+      "loss": 2.0537,
+      "step": 35780
+    },
+    {
+      "epoch": 0.8428712731383354,
+      "grad_norm": 1.8182003498077393,
+      "learning_rate": 3.1496396778295886e-05,
+      "loss": 2.1486,
+      "step": 35790
+    },
+    {
+      "epoch": 0.8431067778248788,
+      "grad_norm": 2.2902069091796875,
+      "learning_rate": 3.1449295840987236e-05,
+      "loss": 2.056,
+      "step": 35800
+    },
+    {
+      "epoch": 0.843342282511422,
+      "grad_norm": 1.7687901258468628,
+      "learning_rate": 3.1402194903678587e-05,
+      "loss": 2.1473,
+      "step": 35810
+    },
+    {
+      "epoch": 0.8435777871979653,
+      "grad_norm": 1.9152394533157349,
+      "learning_rate": 3.135509396636993e-05,
+      "loss": 2.008,
+      "step": 35820
+    },
+    {
+      "epoch": 0.8438132918845085,
+      "grad_norm": 2.308910846710205,
+      "learning_rate": 3.130799302906128e-05,
+      "loss": 1.9928,
+      "step": 35830
+    },
+    {
+      "epoch": 0.8440487965710518,
+      "grad_norm": 1.848122477531433,
+      "learning_rate": 3.1260892091752624e-05,
+      "loss": 2.1119,
+      "step": 35840
+    },
+    {
+      "epoch": 0.844284301257595,
+      "grad_norm": 2.760970115661621,
+      "learning_rate": 3.121379115444398e-05,
+      "loss": 1.9572,
+      "step": 35850
+    },
+    {
+      "epoch": 0.8445198059441383,
+      "grad_norm": 2.340911865234375,
+      "learning_rate": 3.1166690217135324e-05,
+      "loss": 2.0947,
+      "step": 35860
+    },
+    {
+      "epoch": 0.8447553106306815,
+      "grad_norm": 1.8320900201797485,
+      "learning_rate": 3.1119589279826674e-05,
+      "loss": 1.8869,
+      "step": 35870
+    },
+    {
+      "epoch": 0.8449908153172249,
+      "grad_norm": 2.2665231227874756,
+      "learning_rate": 3.107248834251802e-05,
+      "loss": 2.0459,
+      "step": 35880
+    },
+    {
+      "epoch": 0.8452263200037681,
+      "grad_norm": 2.3234121799468994,
+      "learning_rate": 3.102538740520937e-05,
+      "loss": 1.8317,
+      "step": 35890
+    },
+    {
+      "epoch": 0.8454618246903114,
+      "grad_norm": 2.1993918418884277,
+      "learning_rate": 3.097828646790071e-05,
+      "loss": 2.1804,
+      "step": 35900
+    },
+    {
+      "epoch": 0.8456973293768546,
+      "grad_norm": 2.1090517044067383,
+      "learning_rate": 3.093118553059206e-05,
+      "loss": 1.921,
+      "step": 35910
+    },
+    {
+      "epoch": 0.8459328340633978,
+      "grad_norm": 2.0984067916870117,
+      "learning_rate": 3.0884084593283405e-05,
+      "loss": 1.9881,
+      "step": 35920
+    },
+    {
+      "epoch": 0.8461683387499411,
+      "grad_norm": 1.9138286113739014,
+      "learning_rate": 3.0836983655974756e-05,
+      "loss": 1.965,
+      "step": 35930
+    },
+    {
+      "epoch": 0.8464038434364843,
+      "grad_norm": 2.4115593433380127,
+      "learning_rate": 3.07898827186661e-05,
+      "loss": 2.0701,
+      "step": 35940
+    },
+    {
+      "epoch": 0.8466393481230277,
+      "grad_norm": 2.430145025253296,
+      "learning_rate": 3.074278178135745e-05,
+      "loss": 1.8851,
+      "step": 35950
+    },
+    {
+      "epoch": 0.8468748528095709,
+      "grad_norm": 2.130075216293335,
+      "learning_rate": 3.06956808440488e-05,
+      "loss": 2.0154,
+      "step": 35960
+    },
+    {
+      "epoch": 0.8471103574961142,
+      "grad_norm": 2.5573925971984863,
+      "learning_rate": 3.064857990674014e-05,
+      "loss": 2.2392,
+      "step": 35970
+    },
+    {
+      "epoch": 0.8473458621826574,
+      "grad_norm": 1.901427984237671,
+      "learning_rate": 3.060147896943149e-05,
+      "loss": 1.7748,
+      "step": 35980
+    },
+    {
+      "epoch": 0.8475813668692007,
+      "grad_norm": 2.2052485942840576,
+      "learning_rate": 3.0554378032122843e-05,
+      "loss": 1.8585,
+      "step": 35990
+    },
+    {
+      "epoch": 0.8478168715557439,
+      "grad_norm": 2.0382041931152344,
+      "learning_rate": 3.050727709481419e-05,
+      "loss": 1.9512,
+      "step": 36000
+    },
+    {
+      "epoch": 0.8480523762422872,
+      "grad_norm": 2.326741933822632,
+      "learning_rate": 3.0460176157505537e-05,
+      "loss": 2.1158,
+      "step": 36010
+    },
+    {
+      "epoch": 0.8482878809288305,
+      "grad_norm": 2.1949450969696045,
+      "learning_rate": 3.0413075220196884e-05,
+      "loss": 1.7401,
+      "step": 36020
+    },
+    {
+      "epoch": 0.8485233856153738,
+      "grad_norm": 1.9501028060913086,
+      "learning_rate": 3.036597428288823e-05,
+      "loss": 2.1736,
+      "step": 36030
+    },
+    {
+      "epoch": 0.848758890301917,
+      "grad_norm": 1.9012770652770996,
+      "learning_rate": 3.0318873345579578e-05,
+      "loss": 1.9965,
+      "step": 36040
+    },
+    {
+      "epoch": 0.8489943949884603,
+      "grad_norm": 2.4797487258911133,
+      "learning_rate": 3.0271772408270928e-05,
+      "loss": 2.0385,
+      "step": 36050
+    },
+    {
+      "epoch": 0.8492298996750035,
+      "grad_norm": 2.465705394744873,
+      "learning_rate": 3.0224671470962275e-05,
+      "loss": 1.9674,
+      "step": 36060
+    },
+    {
+      "epoch": 0.8494654043615468,
+      "grad_norm": 3.0579581260681152,
+      "learning_rate": 3.0177570533653622e-05,
+      "loss": 1.9021,
+      "step": 36070
+    },
+    {
+      "epoch": 0.84970090904809,
+      "grad_norm": 1.4716150760650635,
+      "learning_rate": 3.013046959634497e-05,
+      "loss": 1.9811,
+      "step": 36080
+    },
+    {
+      "epoch": 0.8499364137346334,
+      "grad_norm": 2.094615936279297,
+      "learning_rate": 3.0083368659036315e-05,
+      "loss": 2.1144,
+      "step": 36090
+    },
+    {
+      "epoch": 0.8501719184211766,
+      "grad_norm": 2.2217042446136475,
+      "learning_rate": 3.0036267721727662e-05,
+      "loss": 2.1761,
+      "step": 36100
+    },
+    {
+      "epoch": 0.8504074231077199,
+      "grad_norm": 2.1508212089538574,
+      "learning_rate": 2.998916678441901e-05,
+      "loss": 2.0179,
+      "step": 36110
+    },
+    {
+      "epoch": 0.8506429277942631,
+      "grad_norm": 2.1316213607788086,
+      "learning_rate": 2.9942065847110356e-05,
+      "loss": 2.0658,
+      "step": 36120
+    },
+    {
+      "epoch": 0.8508784324808064,
+      "grad_norm": 2.383114814758301,
+      "learning_rate": 2.9894964909801703e-05,
+      "loss": 1.9316,
+      "step": 36130
+    },
+    {
+      "epoch": 0.8511139371673496,
+      "grad_norm": 2.28301739692688,
+      "learning_rate": 2.9847863972493056e-05,
+      "loss": 2.1035,
+      "step": 36140
+    },
+    {
+      "epoch": 0.8513494418538929,
+      "grad_norm": 2.0573742389678955,
+      "learning_rate": 2.9800763035184403e-05,
+      "loss": 2.0037,
+      "step": 36150
+    },
+    {
+      "epoch": 0.8515849465404361,
+      "grad_norm": 2.18030047416687,
+      "learning_rate": 2.975366209787575e-05,
+      "loss": 1.925,
+      "step": 36160
+    },
+    {
+      "epoch": 0.8518204512269795,
+      "grad_norm": 2.2723090648651123,
+      "learning_rate": 2.9706561160567097e-05,
+      "loss": 2.1188,
+      "step": 36170
+    },
+    {
+      "epoch": 0.8520559559135227,
+      "grad_norm": 2.1768040657043457,
+      "learning_rate": 2.9659460223258444e-05,
+      "loss": 1.9652,
+      "step": 36180
+    },
+    {
+      "epoch": 0.8522914606000659,
+      "grad_norm": 2.63767671585083,
+      "learning_rate": 2.961235928594979e-05,
+      "loss": 2.0274,
+      "step": 36190
+    },
+    {
+      "epoch": 0.8525269652866092,
+      "grad_norm": 2.380066394805908,
+      "learning_rate": 2.956525834864114e-05,
+      "loss": 1.9089,
+      "step": 36200
+    },
+    {
+      "epoch": 0.8527624699731524,
+      "grad_norm": 1.9944638013839722,
+      "learning_rate": 2.9518157411332488e-05,
+      "loss": 1.815,
+      "step": 36210
+    },
+    {
+      "epoch": 0.8529979746596957,
+      "grad_norm": 1.8857179880142212,
+      "learning_rate": 2.9471056474023835e-05,
+      "loss": 1.8514,
+      "step": 36220
+    },
+    {
+      "epoch": 0.8532334793462389,
+      "grad_norm": 1.8483797311782837,
+      "learning_rate": 2.942395553671518e-05,
+      "loss": 1.9566,
+      "step": 36230
+    },
+    {
+      "epoch": 0.8534689840327823,
+      "grad_norm": 2.12707781791687,
+      "learning_rate": 2.937685459940653e-05,
+      "loss": 2.0217,
+      "step": 36240
+    },
+    {
+      "epoch": 0.8537044887193255,
+      "grad_norm": 2.1420063972473145,
+      "learning_rate": 2.9329753662097875e-05,
+      "loss": 2.2063,
+      "step": 36250
+    },
+    {
+      "epoch": 0.8539399934058688,
+      "grad_norm": 2.08191180229187,
+      "learning_rate": 2.9282652724789222e-05,
+      "loss": 1.9304,
+      "step": 36260
+    },
+    {
+      "epoch": 0.854175498092412,
+      "grad_norm": 1.8427835702896118,
+      "learning_rate": 2.923555178748057e-05,
+      "loss": 1.7183,
+      "step": 36270
+    },
+    {
+      "epoch": 0.8544110027789553,
+      "grad_norm": 2.656949996948242,
+      "learning_rate": 2.9188450850171923e-05,
+      "loss": 2.0153,
+      "step": 36280
+    },
+    {
+      "epoch": 0.8546465074654985,
+      "grad_norm": 1.9904245138168335,
+      "learning_rate": 2.914134991286327e-05,
+      "loss": 1.9786,
+      "step": 36290
+    },
+    {
+      "epoch": 0.8548820121520418,
+      "grad_norm": 2.09130859375,
+      "learning_rate": 2.9094248975554616e-05,
+      "loss": 1.9968,
+      "step": 36300
+    },
+    {
+      "epoch": 0.8551175168385851,
+      "grad_norm": 2.567244529724121,
+      "learning_rate": 2.9047148038245963e-05,
+      "loss": 1.9054,
+      "step": 36310
+    },
+    {
+      "epoch": 0.8553530215251284,
+      "grad_norm": 2.4783241748809814,
+      "learning_rate": 2.900004710093731e-05,
+      "loss": 1.9424,
+      "step": 36320
+    },
+    {
+      "epoch": 0.8555885262116716,
+      "grad_norm": 2.3222262859344482,
+      "learning_rate": 2.8952946163628657e-05,
+      "loss": 2.0085,
+      "step": 36330
+    },
+    {
+      "epoch": 0.8558240308982149,
+      "grad_norm": 2.0765700340270996,
+      "learning_rate": 2.8905845226320004e-05,
+      "loss": 1.9701,
+      "step": 36340
+    },
+    {
+      "epoch": 0.8560595355847581,
+      "grad_norm": 2.526127815246582,
+      "learning_rate": 2.8858744289011354e-05,
+      "loss": 2.0037,
+      "step": 36350
+    },
+    {
+      "epoch": 0.8562950402713014,
+      "grad_norm": 2.3003344535827637,
+      "learning_rate": 2.88116433517027e-05,
+      "loss": 2.0852,
+      "step": 36360
+    },
+    {
+      "epoch": 0.8565305449578446,
+      "grad_norm": 2.28483247756958,
+      "learning_rate": 2.8764542414394048e-05,
+      "loss": 2.1826,
+      "step": 36370
+    },
+    {
+      "epoch": 0.856766049644388,
+      "grad_norm": 2.5736210346221924,
+      "learning_rate": 2.8717441477085395e-05,
+      "loss": 1.9093,
+      "step": 36380
+    },
+    {
+      "epoch": 0.8570015543309312,
+      "grad_norm": 2.280210018157959,
+      "learning_rate": 2.867034053977674e-05,
+      "loss": 2.1804,
+      "step": 36390
+    },
+    {
+      "epoch": 0.8572370590174745,
+      "grad_norm": 2.255749225616455,
+      "learning_rate": 2.8623239602468088e-05,
+      "loss": 2.0126,
+      "step": 36400
+    },
+    {
+      "epoch": 0.8574725637040177,
+      "grad_norm": 2.0187699794769287,
+      "learning_rate": 2.8576138665159435e-05,
+      "loss": 2.2085,
+      "step": 36410
+    },
+    {
+      "epoch": 0.857708068390561,
+      "grad_norm": 3.545119285583496,
+      "learning_rate": 2.852903772785079e-05,
+      "loss": 2.1997,
+      "step": 36420
+    },
+    {
+      "epoch": 0.8579435730771042,
+      "grad_norm": 2.619332790374756,
+      "learning_rate": 2.8481936790542136e-05,
+      "loss": 2.0218,
+      "step": 36430
+    },
+    {
+      "epoch": 0.8581790777636475,
+      "grad_norm": 2.5285348892211914,
+      "learning_rate": 2.8434835853233483e-05,
+      "loss": 1.9183,
+      "step": 36440
+    },
+    {
+      "epoch": 0.8584145824501908,
+      "grad_norm": 3.5454587936401367,
+      "learning_rate": 2.838773491592483e-05,
+      "loss": 2.0172,
+      "step": 36450
+    },
+    {
+      "epoch": 0.8586500871367341,
+      "grad_norm": 1.9408241510391235,
+      "learning_rate": 2.8340633978616176e-05,
+      "loss": 2.2235,
+      "step": 36460
+    },
+    {
+      "epoch": 0.8588855918232773,
+      "grad_norm": 2.6820948123931885,
+      "learning_rate": 2.8293533041307523e-05,
+      "loss": 2.005,
+      "step": 36470
+    },
+    {
+      "epoch": 0.8591210965098205,
+      "grad_norm": 1.9332228899002075,
+      "learning_rate": 2.824643210399887e-05,
+      "loss": 1.9547,
+      "step": 36480
+    },
+    {
+      "epoch": 0.8593566011963638,
+      "grad_norm": 1.865426778793335,
+      "learning_rate": 2.8199331166690217e-05,
+      "loss": 1.8152,
+      "step": 36490
+    },
+    {
+      "epoch": 0.859592105882907,
+      "grad_norm": 1.7702504396438599,
+      "learning_rate": 2.8152230229381567e-05,
+      "loss": 2.0508,
+      "step": 36500
+    },
+    {
+      "epoch": 0.8598276105694503,
+      "grad_norm": 1.7681245803833008,
+      "learning_rate": 2.8105129292072914e-05,
+      "loss": 1.9025,
+      "step": 36510
+    },
+    {
+      "epoch": 0.8600631152559935,
+      "grad_norm": 1.7548736333847046,
+      "learning_rate": 2.805802835476426e-05,
+      "loss": 1.7362,
+      "step": 36520
+    },
+    {
+      "epoch": 0.8602986199425369,
+      "grad_norm": 2.2081425189971924,
+      "learning_rate": 2.8010927417455608e-05,
+      "loss": 2.0412,
+      "step": 36530
+    },
+    {
+      "epoch": 0.8605341246290801,
+      "grad_norm": 1.8118314743041992,
+      "learning_rate": 2.7963826480146954e-05,
+      "loss": 1.9552,
+      "step": 36540
+    },
+    {
+      "epoch": 0.8607696293156234,
+      "grad_norm": 2.4875271320343018,
+      "learning_rate": 2.79167255428383e-05,
+      "loss": 2.0385,
+      "step": 36550
+    },
+    {
+      "epoch": 0.8610051340021666,
+      "grad_norm": 2.542076349258423,
+      "learning_rate": 2.7869624605529655e-05,
+      "loss": 2.1191,
+      "step": 36560
+    },
+    {
+      "epoch": 0.8612406386887099,
+      "grad_norm": 2.1214327812194824,
+      "learning_rate": 2.7822523668221002e-05,
+      "loss": 2.0365,
+      "step": 36570
+    },
+    {
+      "epoch": 0.8614761433752531,
+      "grad_norm": 4.163021564483643,
+      "learning_rate": 2.777542273091235e-05,
+      "loss": 2.0507,
+      "step": 36580
+    },
+    {
+      "epoch": 0.8617116480617965,
+      "grad_norm": 2.633694648742676,
+      "learning_rate": 2.7728321793603696e-05,
+      "loss": 1.9404,
+      "step": 36590
+    },
+    {
+      "epoch": 0.8619471527483397,
+      "grad_norm": 2.0306925773620605,
+      "learning_rate": 2.7681220856295042e-05,
+      "loss": 1.8616,
+      "step": 36600
+    },
+    {
+      "epoch": 0.862182657434883,
+      "grad_norm": 2.5609326362609863,
+      "learning_rate": 2.763411991898639e-05,
+      "loss": 2.0449,
+      "step": 36610
+    },
+    {
+      "epoch": 0.8624181621214262,
+      "grad_norm": 3.2793045043945312,
+      "learning_rate": 2.7587018981677736e-05,
+      "loss": 1.8033,
+      "step": 36620
+    },
+    {
+      "epoch": 0.8626536668079695,
+      "grad_norm": 2.455267906188965,
+      "learning_rate": 2.7539918044369083e-05,
+      "loss": 2.0809,
+      "step": 36630
+    },
+    {
+      "epoch": 0.8628891714945127,
+      "grad_norm": 2.3325164318084717,
+      "learning_rate": 2.749281710706043e-05,
+      "loss": 2.2098,
+      "step": 36640
+    },
+    {
+      "epoch": 0.863124676181056,
+      "grad_norm": 2.911341667175293,
+      "learning_rate": 2.744571616975178e-05,
+      "loss": 1.9998,
+      "step": 36650
+    },
+    {
+      "epoch": 0.8633601808675992,
+      "grad_norm": 1.7594044208526611,
+      "learning_rate": 2.7398615232443127e-05,
+      "loss": 1.9491,
+      "step": 36660
+    },
+    {
+      "epoch": 0.8635956855541426,
+      "grad_norm": 1.8261598348617554,
+      "learning_rate": 2.7351514295134474e-05,
+      "loss": 1.9535,
+      "step": 36670
+    },
+    {
+      "epoch": 0.8638311902406858,
+      "grad_norm": 2.3168840408325195,
+      "learning_rate": 2.730441335782582e-05,
+      "loss": 2.0609,
+      "step": 36680
+    },
+    {
+      "epoch": 0.8640666949272291,
+      "grad_norm": NaN,
+      "learning_rate": 2.7257312420517167e-05,
+      "loss": 2.0912,
+      "step": 36690
+    },
+    {
+      "epoch": 0.8643021996137723,
+      "grad_norm": 2.733128786087036,
+      "learning_rate": 2.7214921576939385e-05,
+      "loss": 1.8064,
+      "step": 36700
+    },
+    {
+      "epoch": 0.8645377043003156,
+      "grad_norm": 2.325484275817871,
+      "learning_rate": 2.7167820639630732e-05,
+      "loss": 2.0417,
+      "step": 36710
+    },
+    {
+      "epoch": 0.8647732089868588,
+      "grad_norm": 2.0244216918945312,
+      "learning_rate": 2.712071970232208e-05,
+      "loss": 1.9551,
+      "step": 36720
+    },
+    {
+      "epoch": 0.8650087136734022,
+      "grad_norm": 3.063344955444336,
+      "learning_rate": 2.7073618765013426e-05,
+      "loss": 2.0306,
+      "step": 36730
+    },
+    {
+      "epoch": 0.8652442183599454,
+      "grad_norm": 2.5917723178863525,
+      "learning_rate": 2.7026517827704773e-05,
+      "loss": 1.9062,
+      "step": 36740
+    },
+    {
+      "epoch": 0.8654797230464887,
+      "grad_norm": 1.9023863077163696,
+      "learning_rate": 2.697941689039612e-05,
+      "loss": 2.0226,
+      "step": 36750
+    },
+    {
+      "epoch": 0.8657152277330319,
+      "grad_norm": 2.4680817127227783,
+      "learning_rate": 2.6932315953087467e-05,
+      "loss": 2.0635,
+      "step": 36760
+    },
+    {
+      "epoch": 0.8659507324195751,
+      "grad_norm": 1.7801569700241089,
+      "learning_rate": 2.6885215015778813e-05,
+      "loss": 1.7529,
+      "step": 36770
+    },
+    {
+      "epoch": 0.8661862371061184,
+      "grad_norm": 2.0391175746917725,
+      "learning_rate": 2.683811407847016e-05,
+      "loss": 2.1877,
+      "step": 36780
+    },
+    {
+      "epoch": 0.8664217417926616,
+      "grad_norm": 1.8722394704818726,
+      "learning_rate": 2.6791013141161507e-05,
+      "loss": 2.0154,
+      "step": 36790
+    },
+    {
+      "epoch": 0.866657246479205,
+      "grad_norm": 1.8816248178482056,
+      "learning_rate": 2.6743912203852854e-05,
+      "loss": 1.9464,
+      "step": 36800
+    },
+    {
+      "epoch": 0.8668927511657482,
+      "grad_norm": 2.1142876148223877,
+      "learning_rate": 2.6696811266544208e-05,
+      "loss": 2.108,
+      "step": 36810
+    },
+    {
+      "epoch": 0.8671282558522915,
+      "grad_norm": 2.861016273498535,
+      "learning_rate": 2.6649710329235554e-05,
+      "loss": 1.998,
+      "step": 36820
+    },
+    {
+      "epoch": 0.8673637605388347,
+      "grad_norm": 2.2118663787841797,
+      "learning_rate": 2.66026093919269e-05,
+      "loss": 2.0853,
+      "step": 36830
+    },
+    {
+      "epoch": 0.867599265225378,
+      "grad_norm": 2.2175252437591553,
+      "learning_rate": 2.6555508454618248e-05,
+      "loss": 1.8626,
+      "step": 36840
+    },
+    {
+      "epoch": 0.8678347699119212,
+      "grad_norm": 1.7482407093048096,
+      "learning_rate": 2.65084075173096e-05,
+      "loss": 2.0537,
+      "step": 36850
+    },
+    {
+      "epoch": 0.8680702745984645,
+      "grad_norm": 1.8986228704452515,
+      "learning_rate": 2.6461306580000945e-05,
+      "loss": 2.0676,
+      "step": 36860
+    },
+    {
+      "epoch": 0.8683057792850077,
+      "grad_norm": 3.3696067333221436,
+      "learning_rate": 2.6414205642692292e-05,
+      "loss": 2.1246,
+      "step": 36870
+    },
+    {
+      "epoch": 0.8685412839715511,
+      "grad_norm": 2.3474841117858887,
+      "learning_rate": 2.636710470538364e-05,
+      "loss": 2.0152,
+      "step": 36880
+    },
+    {
+      "epoch": 0.8687767886580943,
+      "grad_norm": 2.315103769302368,
+      "learning_rate": 2.6320003768074986e-05,
+      "loss": 2.1758,
+      "step": 36890
+    },
+    {
+      "epoch": 0.8690122933446376,
+      "grad_norm": 1.727292776107788,
+      "learning_rate": 2.6272902830766333e-05,
+      "loss": 2.0967,
+      "step": 36900
+    },
+    {
+      "epoch": 0.8692477980311808,
+      "grad_norm": 2.4542670249938965,
+      "learning_rate": 2.622580189345768e-05,
+      "loss": 1.9784,
+      "step": 36910
+    },
+    {
+      "epoch": 0.8694833027177241,
+      "grad_norm": 2.198284149169922,
+      "learning_rate": 2.6178700956149026e-05,
+      "loss": 1.8952,
+      "step": 36920
+    },
+    {
+      "epoch": 0.8697188074042673,
+      "grad_norm": 3.0475144386291504,
+      "learning_rate": 2.6131600018840373e-05,
+      "loss": 1.8998,
+      "step": 36930
+    },
+    {
+      "epoch": 0.8699543120908106,
+      "grad_norm": 1.985602855682373,
+      "learning_rate": 2.608449908153172e-05,
+      "loss": 1.948,
+      "step": 36940
+    },
+    {
+      "epoch": 0.8701898167773539,
+      "grad_norm": 2.757429838180542,
+      "learning_rate": 2.6037398144223067e-05,
+      "loss": 1.9312,
+      "step": 36950
+    },
+    {
+      "epoch": 0.8704253214638972,
+      "grad_norm": 2.614267349243164,
+      "learning_rate": 2.599029720691442e-05,
+      "loss": 2.0093,
+      "step": 36960
+    },
+    {
+      "epoch": 0.8706608261504404,
+      "grad_norm": 1.770949363708496,
+      "learning_rate": 2.5943196269605767e-05,
+      "loss": 1.9344,
+      "step": 36970
+    },
+    {
+      "epoch": 0.8708963308369837,
+      "grad_norm": 2.9338138103485107,
+      "learning_rate": 2.5896095332297114e-05,
+      "loss": 1.9617,
+      "step": 36980
+    },
+    {
+      "epoch": 0.8711318355235269,
+      "grad_norm": 2.102163076400757,
+      "learning_rate": 2.584899439498846e-05,
+      "loss": 2.0294,
+      "step": 36990
+    },
+    {
+      "epoch": 0.8713673402100702,
+      "grad_norm": 1.8639206886291504,
+      "learning_rate": 2.580189345767981e-05,
+      "loss": 1.9084,
+      "step": 37000
+    },
+    {
+      "epoch": 0.8716028448966134,
+      "grad_norm": 2.295996904373169,
+      "learning_rate": 2.5754792520371158e-05,
+      "loss": 2.081,
+      "step": 37010
+    },
+    {
+      "epoch": 0.8718383495831568,
+      "grad_norm": 2.4227380752563477,
+      "learning_rate": 2.5707691583062505e-05,
+      "loss": 2.0755,
+      "step": 37020
+    },
+    {
+      "epoch": 0.8720738542697,
+      "grad_norm": 2.3559842109680176,
+      "learning_rate": 2.5660590645753852e-05,
+      "loss": 1.8186,
+      "step": 37030
+    },
+    {
+      "epoch": 0.8723093589562432,
+      "grad_norm": 2.341525077819824,
+      "learning_rate": 2.56134897084452e-05,
+      "loss": 2.0243,
+      "step": 37040
+    },
+    {
+      "epoch": 0.8725448636427865,
+      "grad_norm": 3.58907413482666,
+      "learning_rate": 2.5566388771136546e-05,
+      "loss": 2.1113,
+      "step": 37050
+    },
+    {
+      "epoch": 0.8727803683293297,
+      "grad_norm": 1.8478801250457764,
+      "learning_rate": 2.5519287833827893e-05,
+      "loss": 1.8545,
+      "step": 37060
+    },
+    {
+      "epoch": 0.873015873015873,
+      "grad_norm": 2.5865366458892822,
+      "learning_rate": 2.547218689651924e-05,
+      "loss": 2.1964,
+      "step": 37070
+    },
+    {
+      "epoch": 0.8732513777024162,
+      "grad_norm": 2.055079460144043,
+      "learning_rate": 2.5425085959210586e-05,
+      "loss": 1.9852,
+      "step": 37080
+    },
+    {
+      "epoch": 0.8734868823889596,
+      "grad_norm": 1.9942408800125122,
+      "learning_rate": 2.5377985021901933e-05,
+      "loss": 2.0406,
+      "step": 37090
+    },
+    {
+      "epoch": 0.8737223870755028,
+      "grad_norm": 2.4995577335357666,
+      "learning_rate": 2.5330884084593287e-05,
+      "loss": 2.0003,
+      "step": 37100
+    },
+    {
+      "epoch": 0.8739578917620461,
+      "grad_norm": 2.066697597503662,
+      "learning_rate": 2.5283783147284634e-05,
+      "loss": 1.9404,
+      "step": 37110
+    },
+    {
+      "epoch": 0.8741933964485893,
+      "grad_norm": 2.102442741394043,
+      "learning_rate": 2.523668220997598e-05,
+      "loss": 1.9775,
+      "step": 37120
+    },
+    {
+      "epoch": 0.8744289011351326,
+      "grad_norm": 2.435946464538574,
+      "learning_rate": 2.5189581272667327e-05,
+      "loss": 2.0967,
+      "step": 37130
+    },
+    {
+      "epoch": 0.8746644058216758,
+      "grad_norm": 1.9978781938552856,
+      "learning_rate": 2.5142480335358674e-05,
+      "loss": 2.0675,
+      "step": 37140
+    },
+    {
+      "epoch": 0.8748999105082191,
+      "grad_norm": 2.7619597911834717,
+      "learning_rate": 2.5095379398050024e-05,
+      "loss": 2.0668,
+      "step": 37150
+    },
+    {
+      "epoch": 0.8751354151947623,
+      "grad_norm": 1.664589285850525,
+      "learning_rate": 2.504827846074137e-05,
+      "loss": 2.1728,
+      "step": 37160
+    },
+    {
+      "epoch": 0.8753709198813057,
+      "grad_norm": 2.1075856685638428,
+      "learning_rate": 2.5001177523432718e-05,
+      "loss": 1.7206,
+      "step": 37170
+    },
+    {
+      "epoch": 0.8756064245678489,
+      "grad_norm": 2.175349473953247,
+      "learning_rate": 2.4954076586124065e-05,
+      "loss": 1.9852,
+      "step": 37180
+    },
+    {
+      "epoch": 0.8758419292543922,
+      "grad_norm": 2.1654629707336426,
+      "learning_rate": 2.4906975648815412e-05,
+      "loss": 1.9823,
+      "step": 37190
+    },
+    {
+      "epoch": 0.8760774339409354,
+      "grad_norm": 2.405961036682129,
+      "learning_rate": 2.4859874711506762e-05,
+      "loss": 1.8534,
+      "step": 37200
+    },
+    {
+      "epoch": 0.8763129386274787,
+      "grad_norm": 1.9819753170013428,
+      "learning_rate": 2.481277377419811e-05,
+      "loss": 2.1591,
+      "step": 37210
+    },
+    {
+      "epoch": 0.8765484433140219,
+      "grad_norm": 2.0572049617767334,
+      "learning_rate": 2.4765672836889456e-05,
+      "loss": 1.9411,
+      "step": 37220
+    },
+    {
+      "epoch": 0.8767839480005652,
+      "grad_norm": 2.2193071842193604,
+      "learning_rate": 2.4718571899580803e-05,
+      "loss": 1.9532,
+      "step": 37230
+    },
+    {
+      "epoch": 0.8770194526871085,
+      "grad_norm": 2.0980918407440186,
+      "learning_rate": 2.467147096227215e-05,
+      "loss": 1.9071,
+      "step": 37240
+    },
+    {
+      "epoch": 0.8772549573736518,
+      "grad_norm": 2.7470831871032715,
+      "learning_rate": 2.4624370024963496e-05,
+      "loss": 2.072,
+      "step": 37250
+    },
+    {
+      "epoch": 0.877490462060195,
+      "grad_norm": 1.7608436346054077,
+      "learning_rate": 2.4577269087654843e-05,
+      "loss": 2.0747,
+      "step": 37260
+    },
+    {
+      "epoch": 0.8777259667467383,
+      "grad_norm": 1.77220618724823,
+      "learning_rate": 2.4530168150346193e-05,
+      "loss": 1.7371,
+      "step": 37270
+    },
+    {
+      "epoch": 0.8779614714332815,
+      "grad_norm": 2.003990411758423,
+      "learning_rate": 2.448306721303754e-05,
+      "loss": 2.0556,
+      "step": 37280
+    },
+    {
+      "epoch": 0.8781969761198248,
+      "grad_norm": 2.2261626720428467,
+      "learning_rate": 2.4435966275728887e-05,
+      "loss": 2.1423,
+      "step": 37290
+    },
+    {
+      "epoch": 0.878432480806368,
+      "grad_norm": 2.8093926906585693,
+      "learning_rate": 2.4388865338420237e-05,
+      "loss": 2.0423,
+      "step": 37300
+    },
+    {
+      "epoch": 0.8786679854929114,
+      "grad_norm": 2.5685336589813232,
+      "learning_rate": 2.4341764401111584e-05,
+      "loss": 2.134,
+      "step": 37310
+    },
+    {
+      "epoch": 0.8789034901794546,
+      "grad_norm": 2.4057536125183105,
+      "learning_rate": 2.429466346380293e-05,
+      "loss": 2.2208,
+      "step": 37320
+    },
+    {
+      "epoch": 0.8791389948659978,
+      "grad_norm": 1.4305607080459595,
+      "learning_rate": 2.4247562526494278e-05,
+      "loss": 2.0589,
+      "step": 37330
+    },
+    {
+      "epoch": 0.8793744995525411,
+      "grad_norm": 2.37786865234375,
+      "learning_rate": 2.4200461589185628e-05,
+      "loss": 1.9452,
+      "step": 37340
+    },
+    {
+      "epoch": 0.8796100042390843,
+      "grad_norm": 1.7371529340744019,
+      "learning_rate": 2.4153360651876975e-05,
+      "loss": 1.9545,
+      "step": 37350
+    },
+    {
+      "epoch": 0.8798455089256276,
+      "grad_norm": 2.432138681411743,
+      "learning_rate": 2.4106259714568322e-05,
+      "loss": 1.9643,
+      "step": 37360
+    },
+    {
+      "epoch": 0.8800810136121708,
+      "grad_norm": 2.241659641265869,
+      "learning_rate": 2.405915877725967e-05,
+      "loss": 1.946,
+      "step": 37370
+    },
+    {
+      "epoch": 0.8803165182987142,
+      "grad_norm": 2.5859363079071045,
+      "learning_rate": 2.4012057839951016e-05,
+      "loss": 2.0878,
+      "step": 37380
+    },
+    {
+      "epoch": 0.8805520229852574,
+      "grad_norm": 2.5846035480499268,
+      "learning_rate": 2.3964956902642363e-05,
+      "loss": 1.9119,
+      "step": 37390
+    },
+    {
+      "epoch": 0.8807875276718007,
+      "grad_norm": 3.664036512374878,
+      "learning_rate": 2.391785596533371e-05,
+      "loss": 1.9178,
+      "step": 37400
+    },
+    {
+      "epoch": 0.8810230323583439,
+      "grad_norm": 2.499345064163208,
+      "learning_rate": 2.3870755028025056e-05,
+      "loss": 2.0054,
+      "step": 37410
+    },
+    {
+      "epoch": 0.8812585370448872,
+      "grad_norm": 1.7014585733413696,
+      "learning_rate": 2.3823654090716406e-05,
+      "loss": 2.0352,
+      "step": 37420
+    },
+    {
+      "epoch": 0.8814940417314304,
+      "grad_norm": 2.9908242225646973,
+      "learning_rate": 2.3776553153407753e-05,
+      "loss": 1.9084,
+      "step": 37430
+    },
+    {
+      "epoch": 0.8817295464179737,
+      "grad_norm": 2.379347324371338,
+      "learning_rate": 2.37294522160991e-05,
+      "loss": 1.9052,
+      "step": 37440
+    },
+    {
+      "epoch": 0.881965051104517,
+      "grad_norm": 1.8825994729995728,
+      "learning_rate": 2.368235127879045e-05,
+      "loss": 2.0933,
+      "step": 37450
+    },
+    {
+      "epoch": 0.8822005557910603,
+      "grad_norm": 2.112344980239868,
+      "learning_rate": 2.3635250341481797e-05,
+      "loss": 1.945,
+      "step": 37460
+    },
+    {
+      "epoch": 0.8824360604776035,
+      "grad_norm": 2.0043137073516846,
+      "learning_rate": 2.3588149404173144e-05,
+      "loss": 2.0133,
+      "step": 37470
+    },
+    {
+      "epoch": 0.8826715651641468,
+      "grad_norm": 2.380242347717285,
+      "learning_rate": 2.354104846686449e-05,
+      "loss": 2.0031,
+      "step": 37480
+    },
+    {
+      "epoch": 0.88290706985069,
+      "grad_norm": 2.028883457183838,
+      "learning_rate": 2.349394752955584e-05,
+      "loss": 1.8129,
+      "step": 37490
+    },
+    {
+      "epoch": 0.8831425745372333,
+      "grad_norm": 2.6352202892303467,
+      "learning_rate": 2.3446846592247188e-05,
+      "loss": 2.1235,
+      "step": 37500
+    },
+    {
+      "epoch": 0.8833780792237765,
+      "grad_norm": 2.017411470413208,
+      "learning_rate": 2.3399745654938535e-05,
+      "loss": 1.8616,
+      "step": 37510
+    },
+    {
+      "epoch": 0.8836135839103199,
+      "grad_norm": 2.081916093826294,
+      "learning_rate": 2.3352644717629882e-05,
+      "loss": 1.8437,
+      "step": 37520
+    },
+    {
+      "epoch": 0.8838490885968631,
+      "grad_norm": 2.4663143157958984,
+      "learning_rate": 2.330554378032123e-05,
+      "loss": 1.7767,
+      "step": 37530
+    },
+    {
+      "epoch": 0.8840845932834064,
+      "grad_norm": 2.4353020191192627,
+      "learning_rate": 2.3258442843012576e-05,
+      "loss": 2.2355,
+      "step": 37540
+    },
+    {
+      "epoch": 0.8843200979699496,
+      "grad_norm": 2.1264777183532715,
+      "learning_rate": 2.3211341905703922e-05,
+      "loss": 2.0394,
+      "step": 37550
+    },
+    {
+      "epoch": 0.8845556026564929,
+      "grad_norm": 2.2275872230529785,
+      "learning_rate": 2.3164240968395273e-05,
+      "loss": 2.2761,
+      "step": 37560
+    },
+    {
+      "epoch": 0.8847911073430361,
+      "grad_norm": 4.1986284255981445,
+      "learning_rate": 2.311714003108662e-05,
+      "loss": 1.9435,
+      "step": 37570
+    },
+    {
+      "epoch": 0.8850266120295794,
+      "grad_norm": 2.330110549926758,
+      "learning_rate": 2.3070039093777966e-05,
+      "loss": 1.7825,
+      "step": 37580
+    },
+    {
+      "epoch": 0.8852621167161226,
+      "grad_norm": 2.903902769088745,
+      "learning_rate": 2.3022938156469313e-05,
+      "loss": 2.0841,
+      "step": 37590
+    },
+    {
+      "epoch": 0.8854976214026659,
+      "grad_norm": 2.5278985500335693,
+      "learning_rate": 2.2975837219160663e-05,
+      "loss": 2.2633,
+      "step": 37600
+    },
+    {
+      "epoch": 0.8857331260892092,
+      "grad_norm": 2.081477165222168,
+      "learning_rate": 2.292873628185201e-05,
+      "loss": 2.0897,
+      "step": 37610
+    },
+    {
+      "epoch": 0.8859686307757524,
+      "grad_norm": 1.9763598442077637,
+      "learning_rate": 2.2881635344543357e-05,
+      "loss": 1.9917,
+      "step": 37620
+    },
+    {
+      "epoch": 0.8862041354622957,
+      "grad_norm": 2.136435031890869,
+      "learning_rate": 2.2834534407234707e-05,
+      "loss": 1.8923,
+      "step": 37630
+    },
+    {
+      "epoch": 0.8864396401488389,
+      "grad_norm": 2.021329402923584,
+      "learning_rate": 2.2787433469926054e-05,
+      "loss": 2.1512,
+      "step": 37640
+    },
+    {
+      "epoch": 0.8866751448353822,
+      "grad_norm": 2.108199119567871,
+      "learning_rate": 2.27403325326174e-05,
+      "loss": 2.1143,
+      "step": 37650
+    },
+    {
+      "epoch": 0.8869106495219254,
+      "grad_norm": 3.078599691390991,
+      "learning_rate": 2.2693231595308748e-05,
+      "loss": 2.0842,
+      "step": 37660
+    },
+    {
+      "epoch": 0.8871461542084688,
+      "grad_norm": 2.235524892807007,
+      "learning_rate": 2.2646130658000095e-05,
+      "loss": 1.945,
+      "step": 37670
+    },
+    {
+      "epoch": 0.887381658895012,
+      "grad_norm": 2.1809022426605225,
+      "learning_rate": 2.259902972069144e-05,
+      "loss": 1.9934,
+      "step": 37680
+    },
+    {
+      "epoch": 0.8876171635815553,
+      "grad_norm": 2.0931341648101807,
+      "learning_rate": 2.255192878338279e-05,
+      "loss": 2.0573,
+      "step": 37690
+    },
+    {
+      "epoch": 0.8878526682680985,
+      "grad_norm": 2.221452236175537,
+      "learning_rate": 2.250482784607414e-05,
+      "loss": 1.8676,
+      "step": 37700
+    },
+    {
+      "epoch": 0.8880881729546418,
+      "grad_norm": 2.1494803428649902,
+      "learning_rate": 2.2457726908765486e-05,
+      "loss": 2.0305,
+      "step": 37710
+    },
+    {
+      "epoch": 0.888323677641185,
+      "grad_norm": 2.0528335571289062,
+      "learning_rate": 2.2410625971456832e-05,
+      "loss": 1.8773,
+      "step": 37720
+    },
+    {
+      "epoch": 0.8885591823277283,
+      "grad_norm": 2.3743772506713867,
+      "learning_rate": 2.236352503414818e-05,
+      "loss": 2.1906,
+      "step": 37730
+    },
+    {
+      "epoch": 0.8887946870142716,
+      "grad_norm": 2.228240966796875,
+      "learning_rate": 2.2316424096839526e-05,
+      "loss": 1.9563,
+      "step": 37740
+    },
+    {
+      "epoch": 0.8890301917008149,
+      "grad_norm": 2.1875791549682617,
+      "learning_rate": 2.2269323159530876e-05,
+      "loss": 1.8761,
+      "step": 37750
+    },
+    {
+      "epoch": 0.8892656963873581,
+      "grad_norm": 3.2131400108337402,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 1.999,
+      "step": 37760
+    },
+    {
+      "epoch": 0.8895012010739014,
+      "grad_norm": 2.3823819160461426,
+      "learning_rate": 2.2175121284913574e-05,
+      "loss": 1.9455,
+      "step": 37770
+    },
+    {
+      "epoch": 0.8897367057604446,
+      "grad_norm": 2.280933141708374,
+      "learning_rate": 2.212802034760492e-05,
+      "loss": 1.9539,
+      "step": 37780
+    },
+    {
+      "epoch": 0.8899722104469879,
+      "grad_norm": 2.0002353191375732,
+      "learning_rate": 2.2080919410296267e-05,
+      "loss": 1.891,
+      "step": 37790
+    },
+    {
+      "epoch": 0.8902077151335311,
+      "grad_norm": 1.8115776777267456,
+      "learning_rate": 2.2033818472987614e-05,
+      "loss": 2.0144,
+      "step": 37800
+    },
+    {
+      "epoch": 0.8904432198200745,
+      "grad_norm": 2.2819020748138428,
+      "learning_rate": 2.198671753567896e-05,
+      "loss": 1.8285,
+      "step": 37810
+    },
+    {
+      "epoch": 0.8906787245066177,
+      "grad_norm": 2.1518003940582275,
+      "learning_rate": 2.1939616598370308e-05,
+      "loss": 1.9067,
+      "step": 37820
+    },
+    {
+      "epoch": 0.890914229193161,
+      "grad_norm": 2.508162498474121,
+      "learning_rate": 2.1892515661061655e-05,
+      "loss": 2.061,
+      "step": 37830
+    },
+    {
+      "epoch": 0.8911497338797042,
+      "grad_norm": 2.1272640228271484,
+      "learning_rate": 2.1845414723753005e-05,
+      "loss": 1.9815,
+      "step": 37840
+    },
+    {
+      "epoch": 0.8913852385662475,
+      "grad_norm": 3.0076818466186523,
+      "learning_rate": 2.1798313786444352e-05,
+      "loss": 2.0412,
+      "step": 37850
+    },
+    {
+      "epoch": 0.8916207432527907,
+      "grad_norm": 2.511852264404297,
+      "learning_rate": 2.17512128491357e-05,
+      "loss": 2.0381,
+      "step": 37860
+    },
+    {
+      "epoch": 0.891856247939334,
+      "grad_norm": 2.035937547683716,
+      "learning_rate": 2.1704111911827045e-05,
+      "loss": 1.9446,
+      "step": 37870
+    },
+    {
+      "epoch": 0.8920917526258773,
+      "grad_norm": 1.9787641763687134,
+      "learning_rate": 2.1657010974518392e-05,
+      "loss": 1.9567,
+      "step": 37880
+    },
+    {
+      "epoch": 0.8923272573124205,
+      "grad_norm": 2.1247246265411377,
+      "learning_rate": 2.160991003720974e-05,
+      "loss": 2.0504,
+      "step": 37890
+    },
+    {
+      "epoch": 0.8925627619989638,
+      "grad_norm": 2.798614740371704,
+      "learning_rate": 2.156280909990109e-05,
+      "loss": 1.9327,
+      "step": 37900
+    },
+    {
+      "epoch": 0.892798266685507,
+      "grad_norm": 2.846179246902466,
+      "learning_rate": 2.1515708162592436e-05,
+      "loss": 2.185,
+      "step": 37910
+    },
+    {
+      "epoch": 0.8930337713720503,
+      "grad_norm": 2.097317934036255,
+      "learning_rate": 2.1468607225283787e-05,
+      "loss": 1.8594,
+      "step": 37920
+    },
+    {
+      "epoch": 0.8932692760585935,
+      "grad_norm": 1.8242942094802856,
+      "learning_rate": 2.1421506287975133e-05,
+      "loss": 2.1531,
+      "step": 37930
+    },
+    {
+      "epoch": 0.8935047807451368,
+      "grad_norm": 2.1491048336029053,
+      "learning_rate": 2.137440535066648e-05,
+      "loss": 1.8671,
+      "step": 37940
+    },
+    {
+      "epoch": 0.89374028543168,
+      "grad_norm": 2.0003557205200195,
+      "learning_rate": 2.1327304413357827e-05,
+      "loss": 2.0186,
+      "step": 37950
+    },
+    {
+      "epoch": 0.8939757901182234,
+      "grad_norm": 2.0959033966064453,
+      "learning_rate": 2.1280203476049174e-05,
+      "loss": 2.1963,
+      "step": 37960
+    },
+    {
+      "epoch": 0.8942112948047666,
+      "grad_norm": 1.9621938467025757,
+      "learning_rate": 2.123310253874052e-05,
+      "loss": 2.3335,
+      "step": 37970
+    },
+    {
+      "epoch": 0.8944467994913099,
+      "grad_norm": 2.0647966861724854,
+      "learning_rate": 2.118600160143187e-05,
+      "loss": 2.1695,
+      "step": 37980
+    },
+    {
+      "epoch": 0.8946823041778531,
+      "grad_norm": 1.7961957454681396,
+      "learning_rate": 2.1138900664123218e-05,
+      "loss": 1.8972,
+      "step": 37990
+    },
+    {
+      "epoch": 0.8949178088643964,
+      "grad_norm": 2.169931411743164,
+      "learning_rate": 2.1091799726814565e-05,
+      "loss": 2.2119,
+      "step": 38000
+    },
+    {
+      "epoch": 0.8951533135509396,
+      "grad_norm": 2.0266969203948975,
+      "learning_rate": 2.104469878950591e-05,
+      "loss": 1.8381,
+      "step": 38010
+    },
+    {
+      "epoch": 0.895388818237483,
+      "grad_norm": 1.9402709007263184,
+      "learning_rate": 2.099759785219726e-05,
+      "loss": 2.0565,
+      "step": 38020
+    },
+    {
+      "epoch": 0.8956243229240262,
+      "grad_norm": 1.9579286575317383,
+      "learning_rate": 2.0950496914888605e-05,
+      "loss": 1.8636,
+      "step": 38030
+    },
+    {
+      "epoch": 0.8958598276105695,
+      "grad_norm": 2.2410125732421875,
+      "learning_rate": 2.0903395977579952e-05,
+      "loss": 2.0196,
+      "step": 38040
+    },
+    {
+      "epoch": 0.8960953322971127,
+      "grad_norm": 2.3416740894317627,
+      "learning_rate": 2.0856295040271302e-05,
+      "loss": 1.8177,
+      "step": 38050
+    },
+    {
+      "epoch": 0.896330836983656,
+      "grad_norm": 1.8137468099594116,
+      "learning_rate": 2.080919410296265e-05,
+      "loss": 1.9998,
+      "step": 38060
+    },
+    {
+      "epoch": 0.8965663416701992,
+      "grad_norm": 2.1465566158294678,
+      "learning_rate": 2.0762093165654e-05,
+      "loss": 1.9154,
+      "step": 38070
+    },
+    {
+      "epoch": 0.8968018463567425,
+      "grad_norm": 2.6183629035949707,
+      "learning_rate": 2.0714992228345346e-05,
+      "loss": 2.0203,
+      "step": 38080
+    },
+    {
+      "epoch": 0.8970373510432857,
+      "grad_norm": 3.3519785404205322,
+      "learning_rate": 2.0667891291036693e-05,
+      "loss": 2.0246,
+      "step": 38090
+    },
+    {
+      "epoch": 0.8972728557298291,
+      "grad_norm": 2.2263829708099365,
+      "learning_rate": 2.062079035372804e-05,
+      "loss": 2.158,
+      "step": 38100
+    },
+    {
+      "epoch": 0.8975083604163723,
+      "grad_norm": 2.257312536239624,
+      "learning_rate": 2.0573689416419387e-05,
+      "loss": 1.8202,
+      "step": 38110
+    },
+    {
+      "epoch": 0.8977438651029156,
+      "grad_norm": 3.83290696144104,
+      "learning_rate": 2.0526588479110737e-05,
+      "loss": 2.1171,
+      "step": 38120
+    },
+    {
+      "epoch": 0.8979793697894588,
+      "grad_norm": 1.9540990591049194,
+      "learning_rate": 2.0479487541802084e-05,
+      "loss": 2.0101,
+      "step": 38130
+    },
+    {
+      "epoch": 0.8982148744760021,
+      "grad_norm": 2.0866284370422363,
+      "learning_rate": 2.043238660449343e-05,
+      "loss": 1.8839,
+      "step": 38140
+    },
+    {
+      "epoch": 0.8984503791625453,
+      "grad_norm": 1.6635223627090454,
+      "learning_rate": 2.0385285667184778e-05,
+      "loss": 2.076,
+      "step": 38150
+    },
+    {
+      "epoch": 0.8986858838490887,
+      "grad_norm": 2.1118903160095215,
+      "learning_rate": 2.0338184729876125e-05,
+      "loss": 2.0658,
+      "step": 38160
+    },
+    {
+      "epoch": 0.8989213885356319,
+      "grad_norm": 1.8942980766296387,
+      "learning_rate": 2.029108379256747e-05,
+      "loss": 2.0003,
+      "step": 38170
+    },
+    {
+      "epoch": 0.8991568932221751,
+      "grad_norm": 2.1378395557403564,
+      "learning_rate": 2.024398285525882e-05,
+      "loss": 2.0265,
+      "step": 38180
+    },
+    {
+      "epoch": 0.8993923979087184,
+      "grad_norm": 2.5827856063842773,
+      "learning_rate": 2.019688191795017e-05,
+      "loss": 1.9765,
+      "step": 38190
+    },
+    {
+      "epoch": 0.8996279025952616,
+      "grad_norm": 2.509906530380249,
+      "learning_rate": 2.0149780980641515e-05,
+      "loss": 1.8868,
+      "step": 38200
+    },
+    {
+      "epoch": 0.8998634072818049,
+      "grad_norm": 2.7832489013671875,
+      "learning_rate": 2.0102680043332862e-05,
+      "loss": 1.9318,
+      "step": 38210
+    },
+    {
+      "epoch": 0.9000989119683481,
+      "grad_norm": 2.0925934314727783,
+      "learning_rate": 2.0055579106024213e-05,
+      "loss": 1.7913,
+      "step": 38220
+    },
+    {
+      "epoch": 0.9003344166548914,
+      "grad_norm": 2.3135039806365967,
+      "learning_rate": 2.000847816871556e-05,
+      "loss": 2.1221,
+      "step": 38230
+    },
+    {
+      "epoch": 0.9005699213414347,
+      "grad_norm": 2.136977195739746,
+      "learning_rate": 1.9961377231406906e-05,
+      "loss": 2.0687,
+      "step": 38240
+    },
+    {
+      "epoch": 0.900805426027978,
+      "grad_norm": 2.6171984672546387,
+      "learning_rate": 1.9914276294098253e-05,
+      "loss": 1.9261,
+      "step": 38250
+    },
+    {
+      "epoch": 0.9010409307145212,
+      "grad_norm": 3.6002085208892822,
+      "learning_rate": 1.9867175356789603e-05,
+      "loss": 2.0192,
+      "step": 38260
+    },
+    {
+      "epoch": 0.9012764354010645,
+      "grad_norm": 1.5538113117218018,
+      "learning_rate": 1.982007441948095e-05,
+      "loss": 2.016,
+      "step": 38270
+    },
+    {
+      "epoch": 0.9015119400876077,
+      "grad_norm": 2.0241286754608154,
+      "learning_rate": 1.9772973482172297e-05,
+      "loss": 1.8286,
+      "step": 38280
+    },
+    {
+      "epoch": 0.901747444774151,
+      "grad_norm": 1.9949616193771362,
+      "learning_rate": 1.9725872544863644e-05,
+      "loss": 1.8788,
+      "step": 38290
+    },
+    {
+      "epoch": 0.9019829494606942,
+      "grad_norm": 2.359792947769165,
+      "learning_rate": 1.967877160755499e-05,
+      "loss": 1.8023,
+      "step": 38300
+    },
+    {
+      "epoch": 0.9022184541472376,
+      "grad_norm": 2.121299982070923,
+      "learning_rate": 1.9631670670246338e-05,
+      "loss": 2.0192,
+      "step": 38310
+    },
+    {
+      "epoch": 0.9024539588337808,
+      "grad_norm": 2.227768659591675,
+      "learning_rate": 1.9584569732937684e-05,
+      "loss": 2.0309,
+      "step": 38320
+    },
+    {
+      "epoch": 0.9026894635203241,
+      "grad_norm": 2.0290210247039795,
+      "learning_rate": 1.9537468795629035e-05,
+      "loss": 2.1192,
+      "step": 38330
+    },
+    {
+      "epoch": 0.9029249682068673,
+      "grad_norm": 2.3042876720428467,
+      "learning_rate": 1.949036785832038e-05,
+      "loss": 2.1374,
+      "step": 38340
+    },
+    {
+      "epoch": 0.9031604728934106,
+      "grad_norm": 2.5515758991241455,
+      "learning_rate": 1.944326692101173e-05,
+      "loss": 2.2315,
+      "step": 38350
+    },
+    {
+      "epoch": 0.9033959775799538,
+      "grad_norm": 2.387439012527466,
+      "learning_rate": 1.9396165983703075e-05,
+      "loss": 2.1361,
+      "step": 38360
+    },
+    {
+      "epoch": 0.9036314822664971,
+      "grad_norm": 1.8408795595169067,
+      "learning_rate": 1.9349065046394426e-05,
+      "loss": 1.9147,
+      "step": 38370
+    },
+    {
+      "epoch": 0.9038669869530404,
+      "grad_norm": 1.412001132965088,
+      "learning_rate": 1.9301964109085772e-05,
+      "loss": 1.9571,
+      "step": 38380
+    },
+    {
+      "epoch": 0.9041024916395837,
+      "grad_norm": 1.8655277490615845,
+      "learning_rate": 1.925486317177712e-05,
+      "loss": 1.9687,
+      "step": 38390
+    },
+    {
+      "epoch": 0.9043379963261269,
+      "grad_norm": 2.4135653972625732,
+      "learning_rate": 1.9207762234468466e-05,
+      "loss": 2.0018,
+      "step": 38400
+    },
+    {
+      "epoch": 0.9045735010126702,
+      "grad_norm": 2.3086419105529785,
+      "learning_rate": 1.9160661297159816e-05,
+      "loss": 2.178,
+      "step": 38410
+    },
+    {
+      "epoch": 0.9048090056992134,
+      "grad_norm": 1.9893290996551514,
+      "learning_rate": 1.9113560359851163e-05,
+      "loss": 2.228,
+      "step": 38420
+    },
+    {
+      "epoch": 0.9050445103857567,
+      "grad_norm": 2.6201305389404297,
+      "learning_rate": 1.906645942254251e-05,
+      "loss": 2.059,
+      "step": 38430
+    },
+    {
+      "epoch": 0.9052800150722999,
+      "grad_norm": 2.3512847423553467,
+      "learning_rate": 1.9019358485233857e-05,
+      "loss": 2.0451,
+      "step": 38440
+    },
+    {
+      "epoch": 0.9055155197588431,
+      "grad_norm": 2.233941078186035,
+      "learning_rate": 1.8972257547925204e-05,
+      "loss": 1.9015,
+      "step": 38450
+    },
+    {
+      "epoch": 0.9057510244453865,
+      "grad_norm": 3.108917713165283,
+      "learning_rate": 1.892515661061655e-05,
+      "loss": 1.979,
+      "step": 38460
+    },
+    {
+      "epoch": 0.9059865291319297,
+      "grad_norm": 2.284076452255249,
+      "learning_rate": 1.8878055673307898e-05,
+      "loss": 1.9659,
+      "step": 38470
+    },
+    {
+      "epoch": 0.906222033818473,
+      "grad_norm": 2.1886470317840576,
+      "learning_rate": 1.8830954735999248e-05,
+      "loss": 2.0916,
+      "step": 38480
+    },
+    {
+      "epoch": 0.9064575385050162,
+      "grad_norm": 1.983109712600708,
+      "learning_rate": 1.8783853798690595e-05,
+      "loss": 2.0744,
+      "step": 38490
+    },
+    {
+      "epoch": 0.9066930431915595,
+      "grad_norm": 1.651294231414795,
+      "learning_rate": 1.873675286138194e-05,
+      "loss": 1.8051,
+      "step": 38500
+    },
+    {
+      "epoch": 0.9069285478781027,
+      "grad_norm": 2.303051710128784,
+      "learning_rate": 1.868965192407329e-05,
+      "loss": 1.7336,
+      "step": 38510
+    },
+    {
+      "epoch": 0.907164052564646,
+      "grad_norm": 1.448872447013855,
+      "learning_rate": 1.864255098676464e-05,
+      "loss": 1.9543,
+      "step": 38520
+    },
+    {
+      "epoch": 0.9073995572511893,
+      "grad_norm": 2.634509563446045,
+      "learning_rate": 1.8595450049455985e-05,
+      "loss": 2.0457,
+      "step": 38530
+    },
+    {
+      "epoch": 0.9076350619377326,
+      "grad_norm": 2.3734359741210938,
+      "learning_rate": 1.8548349112147332e-05,
+      "loss": 2.0179,
+      "step": 38540
+    },
+    {
+      "epoch": 0.9078705666242758,
+      "grad_norm": 2.052739381790161,
+      "learning_rate": 1.8501248174838683e-05,
+      "loss": 2.1796,
+      "step": 38550
+    },
+    {
+      "epoch": 0.9081060713108191,
+      "grad_norm": 2.4247121810913086,
+      "learning_rate": 1.845414723753003e-05,
+      "loss": 2.0285,
+      "step": 38560
+    },
+    {
+      "epoch": 0.9083415759973623,
+      "grad_norm": 2.3661036491394043,
+      "learning_rate": 1.8407046300221376e-05,
+      "loss": 2.1629,
+      "step": 38570
+    },
+    {
+      "epoch": 0.9085770806839056,
+      "grad_norm": 1.7307307720184326,
+      "learning_rate": 1.8359945362912723e-05,
+      "loss": 2.0182,
+      "step": 38580
+    },
+    {
+      "epoch": 0.9088125853704488,
+      "grad_norm": 2.318981885910034,
+      "learning_rate": 1.831284442560407e-05,
+      "loss": 2.031,
+      "step": 38590
+    },
+    {
+      "epoch": 0.9090480900569922,
+      "grad_norm": 2.0078811645507812,
+      "learning_rate": 1.8265743488295417e-05,
+      "loss": 1.8833,
+      "step": 38600
+    },
+    {
+      "epoch": 0.9092835947435354,
+      "grad_norm": 1.9469083547592163,
+      "learning_rate": 1.8218642550986764e-05,
+      "loss": 2.0904,
+      "step": 38610
+    },
+    {
+      "epoch": 0.9095190994300787,
+      "grad_norm": 2.0202481746673584,
+      "learning_rate": 1.8171541613678114e-05,
+      "loss": 1.8596,
+      "step": 38620
+    },
+    {
+      "epoch": 0.9097546041166219,
+      "grad_norm": 2.1429693698883057,
+      "learning_rate": 1.812444067636946e-05,
+      "loss": 1.774,
+      "step": 38630
+    },
+    {
+      "epoch": 0.9099901088031652,
+      "grad_norm": 2.0817933082580566,
+      "learning_rate": 1.8077339739060808e-05,
+      "loss": 2.0136,
+      "step": 38640
+    },
+    {
+      "epoch": 0.9102256134897084,
+      "grad_norm": 2.4810056686401367,
+      "learning_rate": 1.8030238801752154e-05,
+      "loss": 2.0795,
+      "step": 38650
+    },
+    {
+      "epoch": 0.9104611181762517,
+      "grad_norm": 2.75999116897583,
+      "learning_rate": 1.79831378644435e-05,
+      "loss": 2.0032,
+      "step": 38660
+    },
+    {
+      "epoch": 0.910696622862795,
+      "grad_norm": 2.1488144397735596,
+      "learning_rate": 1.793603692713485e-05,
+      "loss": 2.3905,
+      "step": 38670
+    },
+    {
+      "epoch": 0.9109321275493383,
+      "grad_norm": 2.028859853744507,
+      "learning_rate": 1.78889359898262e-05,
+      "loss": 2.0345,
+      "step": 38680
+    },
+    {
+      "epoch": 0.9111676322358815,
+      "grad_norm": 1.8787860870361328,
+      "learning_rate": 1.784183505251755e-05,
+      "loss": 1.9495,
+      "step": 38690
+    },
+    {
+      "epoch": 0.9114031369224248,
+      "grad_norm": 2.0410072803497314,
+      "learning_rate": 1.7794734115208896e-05,
+      "loss": 2.0186,
+      "step": 38700
+    },
+    {
+      "epoch": 0.911638641608968,
+      "grad_norm": 2.9783804416656494,
+      "learning_rate": 1.7747633177900242e-05,
+      "loss": 1.8578,
+      "step": 38710
+    },
+    {
+      "epoch": 0.9118741462955113,
+      "grad_norm": 2.104757785797119,
+      "learning_rate": 1.770053224059159e-05,
+      "loss": 1.8201,
+      "step": 38720
+    },
+    {
+      "epoch": 0.9121096509820545,
+      "grad_norm": 1.9207528829574585,
+      "learning_rate": 1.7653431303282936e-05,
+      "loss": 1.8914,
+      "step": 38730
+    },
+    {
+      "epoch": 0.9123451556685978,
+      "grad_norm": 2.920927047729492,
+      "learning_rate": 1.7606330365974283e-05,
+      "loss": 1.9848,
+      "step": 38740
+    },
+    {
+      "epoch": 0.9125806603551411,
+      "grad_norm": 2.510042190551758,
+      "learning_rate": 1.755922942866563e-05,
+      "loss": 2.1576,
+      "step": 38750
+    },
+    {
+      "epoch": 0.9128161650416843,
+      "grad_norm": 2.3450679779052734,
+      "learning_rate": 1.751212849135698e-05,
+      "loss": 2.2116,
+      "step": 38760
+    },
+    {
+      "epoch": 0.9130516697282276,
+      "grad_norm": 1.95484459400177,
+      "learning_rate": 1.7465027554048327e-05,
+      "loss": 1.9605,
+      "step": 38770
+    },
+    {
+      "epoch": 0.9132871744147708,
+      "grad_norm": 2.005983352661133,
+      "learning_rate": 1.7417926616739674e-05,
+      "loss": 2.0368,
+      "step": 38780
+    },
+    {
+      "epoch": 0.9135226791013141,
+      "grad_norm": 2.1589179039001465,
+      "learning_rate": 1.737082567943102e-05,
+      "loss": 1.906,
+      "step": 38790
+    },
+    {
+      "epoch": 0.9137581837878573,
+      "grad_norm": 1.929579257965088,
+      "learning_rate": 1.7323724742122367e-05,
+      "loss": 2.0555,
+      "step": 38800
+    },
+    {
+      "epoch": 0.9139936884744007,
+      "grad_norm": 2.2730865478515625,
+      "learning_rate": 1.7276623804813714e-05,
+      "loss": 2.1307,
+      "step": 38810
+    },
+    {
+      "epoch": 0.9142291931609439,
+      "grad_norm": 1.4058763980865479,
+      "learning_rate": 1.7229522867505065e-05,
+      "loss": 1.6729,
+      "step": 38820
+    },
+    {
+      "epoch": 0.9144646978474872,
+      "grad_norm": 2.0500447750091553,
+      "learning_rate": 1.718242193019641e-05,
+      "loss": 1.8054,
+      "step": 38830
+    },
+    {
+      "epoch": 0.9147002025340304,
+      "grad_norm": 2.374552011489868,
+      "learning_rate": 1.713532099288776e-05,
+      "loss": 1.8979,
+      "step": 38840
+    },
+    {
+      "epoch": 0.9149357072205737,
+      "grad_norm": 2.021498441696167,
+      "learning_rate": 1.708822005557911e-05,
+      "loss": 2.1172,
+      "step": 38850
+    },
+    {
+      "epoch": 0.9151712119071169,
+      "grad_norm": 2.2136051654815674,
+      "learning_rate": 1.7041119118270455e-05,
+      "loss": 2.0001,
+      "step": 38860
+    },
+    {
+      "epoch": 0.9154067165936602,
+      "grad_norm": 2.6992013454437256,
+      "learning_rate": 1.6994018180961802e-05,
+      "loss": 1.8245,
+      "step": 38870
+    },
+    {
+      "epoch": 0.9156422212802034,
+      "grad_norm": 2.2743868827819824,
+      "learning_rate": 1.694691724365315e-05,
+      "loss": 1.8511,
+      "step": 38880
+    },
+    {
+      "epoch": 0.9158777259667468,
+      "grad_norm": 2.4155523777008057,
+      "learning_rate": 1.6899816306344496e-05,
+      "loss": 1.9802,
+      "step": 38890
+    },
+    {
+      "epoch": 0.91611323065329,
+      "grad_norm": 1.777964472770691,
+      "learning_rate": 1.6852715369035846e-05,
+      "loss": 2.0244,
+      "step": 38900
+    },
+    {
+      "epoch": 0.9163487353398333,
+      "grad_norm": 2.0305655002593994,
+      "learning_rate": 1.6805614431727193e-05,
+      "loss": 1.9335,
+      "step": 38910
+    },
+    {
+      "epoch": 0.9165842400263765,
+      "grad_norm": 2.5534350872039795,
+      "learning_rate": 1.675851349441854e-05,
+      "loss": 2.0425,
+      "step": 38920
+    },
+    {
+      "epoch": 0.9168197447129198,
+      "grad_norm": 1.9499086141586304,
+      "learning_rate": 1.6711412557109887e-05,
+      "loss": 2.0643,
+      "step": 38930
+    },
+    {
+      "epoch": 0.917055249399463,
+      "grad_norm": 2.4477667808532715,
+      "learning_rate": 1.6664311619801234e-05,
+      "loss": 1.8278,
+      "step": 38940
+    },
+    {
+      "epoch": 0.9172907540860064,
+      "grad_norm": 2.7401411533355713,
+      "learning_rate": 1.661721068249258e-05,
+      "loss": 2.1896,
+      "step": 38950
+    },
+    {
+      "epoch": 0.9175262587725496,
+      "grad_norm": 2.0226340293884277,
+      "learning_rate": 1.6570109745183927e-05,
+      "loss": 1.9066,
+      "step": 38960
+    },
+    {
+      "epoch": 0.9177617634590929,
+      "grad_norm": 2.22379469871521,
+      "learning_rate": 1.6523008807875278e-05,
+      "loss": 2.0999,
+      "step": 38970
+    },
+    {
+      "epoch": 0.9179972681456361,
+      "grad_norm": 2.1687543392181396,
+      "learning_rate": 1.6475907870566624e-05,
+      "loss": 2.0633,
+      "step": 38980
+    },
+    {
+      "epoch": 0.9182327728321794,
+      "grad_norm": 3.160351514816284,
+      "learning_rate": 1.6428806933257975e-05,
+      "loss": 2.1236,
+      "step": 38990
+    },
+    {
+      "epoch": 0.9184682775187226,
+      "grad_norm": 2.306687355041504,
+      "learning_rate": 1.638170599594932e-05,
+      "loss": 2.1413,
+      "step": 39000
+    },
+    {
+      "epoch": 0.9187037822052659,
+      "grad_norm": 2.567837953567505,
+      "learning_rate": 1.633460505864067e-05,
+      "loss": 1.8291,
+      "step": 39010
+    },
+    {
+      "epoch": 0.9189392868918091,
+      "grad_norm": 2.327446937561035,
+      "learning_rate": 1.6287504121332015e-05,
+      "loss": 2.0234,
+      "step": 39020
+    },
+    {
+      "epoch": 0.9191747915783524,
+      "grad_norm": 2.052375555038452,
+      "learning_rate": 1.6240403184023362e-05,
+      "loss": 1.9689,
+      "step": 39030
+    },
+    {
+      "epoch": 0.9194102962648957,
+      "grad_norm": 1.6674343347549438,
+      "learning_rate": 1.6193302246714712e-05,
+      "loss": 1.9624,
+      "step": 39040
+    },
+    {
+      "epoch": 0.9196458009514389,
+      "grad_norm": 2.403968572616577,
+      "learning_rate": 1.614620130940606e-05,
+      "loss": 2.0186,
+      "step": 39050
+    },
+    {
+      "epoch": 0.9198813056379822,
+      "grad_norm": 2.198310613632202,
+      "learning_rate": 1.6099100372097406e-05,
+      "loss": 1.9116,
+      "step": 39060
+    },
+    {
+      "epoch": 0.9201168103245254,
+      "grad_norm": 1.7685959339141846,
+      "learning_rate": 1.6051999434788753e-05,
+      "loss": 1.7279,
+      "step": 39070
+    },
+    {
+      "epoch": 0.9203523150110687,
+      "grad_norm": 2.2193026542663574,
+      "learning_rate": 1.60048984974801e-05,
+      "loss": 2.064,
+      "step": 39080
+    },
+    {
+      "epoch": 0.9205878196976119,
+      "grad_norm": 2.313854455947876,
+      "learning_rate": 1.5957797560171447e-05,
+      "loss": 1.9564,
+      "step": 39090
+    },
+    {
+      "epoch": 0.9208233243841553,
+      "grad_norm": 3.075056314468384,
+      "learning_rate": 1.5910696622862793e-05,
+      "loss": 1.8532,
+      "step": 39100
+    },
+    {
+      "epoch": 0.9210588290706985,
+      "grad_norm": 2.181581974029541,
+      "learning_rate": 1.5863595685554144e-05,
+      "loss": 2.0414,
+      "step": 39110
+    },
+    {
+      "epoch": 0.9212943337572418,
+      "grad_norm": 1.9923847913742065,
+      "learning_rate": 1.581649474824549e-05,
+      "loss": 1.8529,
+      "step": 39120
+    },
+    {
+      "epoch": 0.921529838443785,
+      "grad_norm": 2.1557722091674805,
+      "learning_rate": 1.5769393810936837e-05,
+      "loss": 1.8607,
+      "step": 39130
+    },
+    {
+      "epoch": 0.9217653431303283,
+      "grad_norm": 1.7339023351669312,
+      "learning_rate": 1.5722292873628188e-05,
+      "loss": 1.9141,
+      "step": 39140
+    },
+    {
+      "epoch": 0.9220008478168715,
+      "grad_norm": 1.983236312866211,
+      "learning_rate": 1.5675191936319535e-05,
+      "loss": 2.1342,
+      "step": 39150
+    },
+    {
+      "epoch": 0.9222363525034148,
+      "grad_norm": 2.0628926753997803,
+      "learning_rate": 1.562809099901088e-05,
+      "loss": 2.0345,
+      "step": 39160
+    },
+    {
+      "epoch": 0.922471857189958,
+      "grad_norm": 2.1119203567504883,
+      "learning_rate": 1.5580990061702228e-05,
+      "loss": 2.033,
+      "step": 39170
+    },
+    {
+      "epoch": 0.9227073618765014,
+      "grad_norm": 2.4918603897094727,
+      "learning_rate": 1.553388912439358e-05,
+      "loss": 1.9512,
+      "step": 39180
+    },
+    {
+      "epoch": 0.9229428665630446,
+      "grad_norm": 1.8734686374664307,
+      "learning_rate": 1.5486788187084925e-05,
+      "loss": 1.8077,
+      "step": 39190
+    },
+    {
+      "epoch": 0.9231783712495879,
+      "grad_norm": 1.4692710638046265,
+      "learning_rate": 1.5439687249776272e-05,
+      "loss": 2.012,
+      "step": 39200
+    },
+    {
+      "epoch": 0.9234138759361311,
+      "grad_norm": 1.7556036710739136,
+      "learning_rate": 1.539258631246762e-05,
+      "loss": 2.0538,
+      "step": 39210
+    },
+    {
+      "epoch": 0.9236493806226744,
+      "grad_norm": 1.9275639057159424,
+      "learning_rate": 1.5345485375158966e-05,
+      "loss": 2.2427,
+      "step": 39220
+    },
+    {
+      "epoch": 0.9238848853092176,
+      "grad_norm": 2.047739267349243,
+      "learning_rate": 1.5298384437850313e-05,
+      "loss": 1.9106,
+      "step": 39230
+    },
+    {
+      "epoch": 0.924120389995761,
+      "grad_norm": 2.382171630859375,
+      "learning_rate": 1.525128350054166e-05,
+      "loss": 2.0102,
+      "step": 39240
+    },
+    {
+      "epoch": 0.9243558946823042,
+      "grad_norm": 1.8619623184204102,
+      "learning_rate": 1.520418256323301e-05,
+      "loss": 2.1269,
+      "step": 39250
+    },
+    {
+      "epoch": 0.9245913993688475,
+      "grad_norm": 2.0923447608947754,
+      "learning_rate": 1.5157081625924357e-05,
+      "loss": 2.0977,
+      "step": 39260
+    },
+    {
+      "epoch": 0.9248269040553907,
+      "grad_norm": 1.8551141023635864,
+      "learning_rate": 1.5109980688615705e-05,
+      "loss": 1.8932,
+      "step": 39270
+    },
+    {
+      "epoch": 0.925062408741934,
+      "grad_norm": 2.1643242835998535,
+      "learning_rate": 1.5062879751307052e-05,
+      "loss": 2.0233,
+      "step": 39280
+    },
+    {
+      "epoch": 0.9252979134284772,
+      "grad_norm": 1.629560947418213,
+      "learning_rate": 1.5015778813998399e-05,
+      "loss": 1.8223,
+      "step": 39290
+    },
+    {
+      "epoch": 0.9255334181150204,
+      "grad_norm": 1.8448771238327026,
+      "learning_rate": 1.4968677876689746e-05,
+      "loss": 2.0332,
+      "step": 39300
+    },
+    {
+      "epoch": 0.9257689228015638,
+      "grad_norm": 2.0944831371307373,
+      "learning_rate": 1.4921576939381093e-05,
+      "loss": 1.8888,
+      "step": 39310
+    },
+    {
+      "epoch": 0.926004427488107,
+      "grad_norm": 2.4087986946105957,
+      "learning_rate": 1.4874476002072443e-05,
+      "loss": 2.0431,
+      "step": 39320
+    },
+    {
+      "epoch": 0.9262399321746503,
+      "grad_norm": 2.948134422302246,
+      "learning_rate": 1.482737506476379e-05,
+      "loss": 1.7793,
+      "step": 39330
+    },
+    {
+      "epoch": 0.9264754368611935,
+      "grad_norm": 1.8383265733718872,
+      "learning_rate": 1.4780274127455138e-05,
+      "loss": 2.0095,
+      "step": 39340
+    },
+    {
+      "epoch": 0.9267109415477368,
+      "grad_norm": 1.9835010766983032,
+      "learning_rate": 1.4733173190146485e-05,
+      "loss": 2.0252,
+      "step": 39350
+    },
+    {
+      "epoch": 0.92694644623428,
+      "grad_norm": 2.6888811588287354,
+      "learning_rate": 1.4686072252837832e-05,
+      "loss": 2.1416,
+      "step": 39360
+    },
+    {
+      "epoch": 0.9271819509208233,
+      "grad_norm": 2.017975330352783,
+      "learning_rate": 1.4638971315529179e-05,
+      "loss": 1.879,
+      "step": 39370
+    },
+    {
+      "epoch": 0.9274174556073665,
+      "grad_norm": 2.2227847576141357,
+      "learning_rate": 1.4591870378220526e-05,
+      "loss": 2.1047,
+      "step": 39380
+    },
+    {
+      "epoch": 0.9276529602939099,
+      "grad_norm": 1.885798454284668,
+      "learning_rate": 1.4544769440911876e-05,
+      "loss": 1.9589,
+      "step": 39390
+    },
+    {
+      "epoch": 0.9278884649804531,
+      "grad_norm": 1.9954543113708496,
+      "learning_rate": 1.4497668503603223e-05,
+      "loss": 2.0047,
+      "step": 39400
+    },
+    {
+      "epoch": 0.9281239696669964,
+      "grad_norm": 2.5150036811828613,
+      "learning_rate": 1.445056756629457e-05,
+      "loss": 2.0924,
+      "step": 39410
+    },
+    {
+      "epoch": 0.9283594743535396,
+      "grad_norm": 2.4229278564453125,
+      "learning_rate": 1.4403466628985918e-05,
+      "loss": 2.015,
+      "step": 39420
+    },
+    {
+      "epoch": 0.9285949790400829,
+      "grad_norm": 2.048858404159546,
+      "learning_rate": 1.4356365691677265e-05,
+      "loss": 1.9736,
+      "step": 39430
+    },
+    {
+      "epoch": 0.9288304837266261,
+      "grad_norm": 2.147585868835449,
+      "learning_rate": 1.4309264754368612e-05,
+      "loss": 1.9895,
+      "step": 39440
+    },
+    {
+      "epoch": 0.9290659884131695,
+      "grad_norm": 1.9288424253463745,
+      "learning_rate": 1.4262163817059959e-05,
+      "loss": 2.0262,
+      "step": 39450
+    },
+    {
+      "epoch": 0.9293014930997127,
+      "grad_norm": 1.9802799224853516,
+      "learning_rate": 1.4215062879751306e-05,
+      "loss": 1.9161,
+      "step": 39460
+    },
+    {
+      "epoch": 0.929536997786256,
+      "grad_norm": 2.0962471961975098,
+      "learning_rate": 1.4167961942442656e-05,
+      "loss": 1.9732,
+      "step": 39470
+    },
+    {
+      "epoch": 0.9297725024727992,
+      "grad_norm": 2.103632688522339,
+      "learning_rate": 1.4120861005134003e-05,
+      "loss": 1.8515,
+      "step": 39480
+    },
+    {
+      "epoch": 0.9300080071593425,
+      "grad_norm": 2.154771327972412,
+      "learning_rate": 1.4073760067825351e-05,
+      "loss": 1.7922,
+      "step": 39490
+    },
+    {
+      "epoch": 0.9302435118458857,
+      "grad_norm": 2.3538451194763184,
+      "learning_rate": 1.4026659130516698e-05,
+      "loss": 2.0359,
+      "step": 39500
+    },
+    {
+      "epoch": 0.930479016532429,
+      "grad_norm": 2.0333411693573,
+      "learning_rate": 1.3979558193208045e-05,
+      "loss": 2.1012,
+      "step": 39510
+    },
+    {
+      "epoch": 0.9307145212189722,
+      "grad_norm": 2.59187388420105,
+      "learning_rate": 1.3932457255899392e-05,
+      "loss": 2.2647,
+      "step": 39520
+    },
+    {
+      "epoch": 0.9309500259055156,
+      "grad_norm": 2.3052356243133545,
+      "learning_rate": 1.3885356318590739e-05,
+      "loss": 2.0736,
+      "step": 39530
+    },
+    {
+      "epoch": 0.9311855305920588,
+      "grad_norm": 2.131145477294922,
+      "learning_rate": 1.3838255381282089e-05,
+      "loss": 2.0845,
+      "step": 39540
+    },
+    {
+      "epoch": 0.9314210352786021,
+      "grad_norm": 2.3522706031799316,
+      "learning_rate": 1.3791154443973436e-05,
+      "loss": 1.9605,
+      "step": 39550
+    },
+    {
+      "epoch": 0.9316565399651453,
+      "grad_norm": 2.067479372024536,
+      "learning_rate": 1.3744053506664783e-05,
+      "loss": 2.138,
+      "step": 39560
+    },
+    {
+      "epoch": 0.9318920446516886,
+      "grad_norm": 1.7099672555923462,
+      "learning_rate": 1.3696952569356131e-05,
+      "loss": 1.7891,
+      "step": 39570
+    },
+    {
+      "epoch": 0.9321275493382318,
+      "grad_norm": 2.0473127365112305,
+      "learning_rate": 1.3649851632047478e-05,
+      "loss": 2.1291,
+      "step": 39580
+    },
+    {
+      "epoch": 0.932363054024775,
+      "grad_norm": 2.2088100910186768,
+      "learning_rate": 1.3602750694738825e-05,
+      "loss": 1.9411,
+      "step": 39590
+    },
+    {
+      "epoch": 0.9325985587113184,
+      "grad_norm": 2.0423855781555176,
+      "learning_rate": 1.3555649757430172e-05,
+      "loss": 2.0322,
+      "step": 39600
+    },
+    {
+      "epoch": 0.9328340633978616,
+      "grad_norm": 3.2174267768859863,
+      "learning_rate": 1.3508548820121522e-05,
+      "loss": 2.0514,
+      "step": 39610
+    },
+    {
+      "epoch": 0.9330695680844049,
+      "grad_norm": 1.992333173751831,
+      "learning_rate": 1.3461447882812869e-05,
+      "loss": 1.7315,
+      "step": 39620
+    },
+    {
+      "epoch": 0.9333050727709481,
+      "grad_norm": 1.7157886028289795,
+      "learning_rate": 1.3414346945504216e-05,
+      "loss": 2.0937,
+      "step": 39630
+    },
+    {
+      "epoch": 0.9335405774574914,
+      "grad_norm": 2.721675157546997,
+      "learning_rate": 1.3367246008195564e-05,
+      "loss": 2.0178,
+      "step": 39640
+    },
+    {
+      "epoch": 0.9337760821440346,
+      "grad_norm": 2.151686429977417,
+      "learning_rate": 1.3320145070886911e-05,
+      "loss": 2.0767,
+      "step": 39650
+    },
+    {
+      "epoch": 0.934011586830578,
+      "grad_norm": 2.0822360515594482,
+      "learning_rate": 1.3273044133578258e-05,
+      "loss": 1.8964,
+      "step": 39660
+    },
+    {
+      "epoch": 0.9342470915171212,
+      "grad_norm": 2.0849602222442627,
+      "learning_rate": 1.3225943196269605e-05,
+      "loss": 2.074,
+      "step": 39670
+    },
+    {
+      "epoch": 0.9344825962036645,
+      "grad_norm": 1.9029971361160278,
+      "learning_rate": 1.3178842258960955e-05,
+      "loss": 1.7693,
+      "step": 39680
+    },
+    {
+      "epoch": 0.9347181008902077,
+      "grad_norm": 1.9549709558486938,
+      "learning_rate": 1.3131741321652302e-05,
+      "loss": 1.866,
+      "step": 39690
+    },
+    {
+      "epoch": 0.934953605576751,
+      "grad_norm": 3.2078988552093506,
+      "learning_rate": 1.3084640384343649e-05,
+      "loss": 2.0639,
+      "step": 39700
+    },
+    {
+      "epoch": 0.9351891102632942,
+      "grad_norm": 2.0642852783203125,
+      "learning_rate": 1.3037539447034996e-05,
+      "loss": 2.2178,
+      "step": 39710
+    },
+    {
+      "epoch": 0.9354246149498375,
+      "grad_norm": 2.6901581287384033,
+      "learning_rate": 1.2990438509726344e-05,
+      "loss": 1.9731,
+      "step": 39720
+    },
+    {
+      "epoch": 0.9356601196363807,
+      "grad_norm": 2.5783872604370117,
+      "learning_rate": 1.2943337572417691e-05,
+      "loss": 2.009,
+      "step": 39730
+    },
+    {
+      "epoch": 0.9358956243229241,
+      "grad_norm": 2.1882944107055664,
+      "learning_rate": 1.2896236635109038e-05,
+      "loss": 1.7685,
+      "step": 39740
+    },
+    {
+      "epoch": 0.9361311290094673,
+      "grad_norm": 2.3130943775177,
+      "learning_rate": 1.2849135697800388e-05,
+      "loss": 1.8615,
+      "step": 39750
+    },
+    {
+      "epoch": 0.9363666336960106,
+      "grad_norm": 2.8880362510681152,
+      "learning_rate": 1.2802034760491735e-05,
+      "loss": 1.6732,
+      "step": 39760
+    },
+    {
+      "epoch": 0.9366021383825538,
+      "grad_norm": 2.318843364715576,
+      "learning_rate": 1.2754933823183082e-05,
+      "loss": 1.7798,
+      "step": 39770
+    },
+    {
+      "epoch": 0.9368376430690971,
+      "grad_norm": 2.4206154346466064,
+      "learning_rate": 1.2707832885874429e-05,
+      "loss": 1.8153,
+      "step": 39780
+    },
+    {
+      "epoch": 0.9370731477556403,
+      "grad_norm": 2.1725382804870605,
+      "learning_rate": 1.2660731948565777e-05,
+      "loss": 1.8012,
+      "step": 39790
+    },
+    {
+      "epoch": 0.9373086524421836,
+      "grad_norm": 2.6982181072235107,
+      "learning_rate": 1.2613631011257124e-05,
+      "loss": 1.9887,
+      "step": 39800
+    },
+    {
+      "epoch": 0.9375441571287269,
+      "grad_norm": 2.1166913509368896,
+      "learning_rate": 1.2566530073948471e-05,
+      "loss": 1.793,
+      "step": 39810
+    },
+    {
+      "epoch": 0.9377796618152702,
+      "grad_norm": 2.248126745223999,
+      "learning_rate": 1.2519429136639821e-05,
+      "loss": 1.9725,
+      "step": 39820
+    },
+    {
+      "epoch": 0.9380151665018134,
+      "grad_norm": 2.303269386291504,
+      "learning_rate": 1.2472328199331166e-05,
+      "loss": 1.9097,
+      "step": 39830
+    },
+    {
+      "epoch": 0.9382506711883567,
+      "grad_norm": 2.438401460647583,
+      "learning_rate": 1.2425227262022515e-05,
+      "loss": 1.8483,
+      "step": 39840
+    },
+    {
+      "epoch": 0.9384861758748999,
+      "grad_norm": 1.9419487714767456,
+      "learning_rate": 1.2378126324713862e-05,
+      "loss": 1.9396,
+      "step": 39850
+    },
+    {
+      "epoch": 0.9387216805614432,
+      "grad_norm": 2.5072529315948486,
+      "learning_rate": 1.2331025387405209e-05,
+      "loss": 1.9367,
+      "step": 39860
+    },
+    {
+      "epoch": 0.9389571852479864,
+      "grad_norm": 2.39945387840271,
+      "learning_rate": 1.2283924450096557e-05,
+      "loss": 1.8372,
+      "step": 39870
+    },
+    {
+      "epoch": 0.9391926899345296,
+      "grad_norm": 2.786134958267212,
+      "learning_rate": 1.2236823512787906e-05,
+      "loss": 1.9416,
+      "step": 39880
+    },
+    {
+      "epoch": 0.939428194621073,
+      "grad_norm": 2.2066538333892822,
+      "learning_rate": 1.2189722575479253e-05,
+      "loss": 2.0354,
+      "step": 39890
+    },
+    {
+      "epoch": 0.9396636993076162,
+      "grad_norm": 2.544435739517212,
+      "learning_rate": 1.21426216381706e-05,
+      "loss": 2.005,
+      "step": 39900
+    },
+    {
+      "epoch": 0.9398992039941595,
+      "grad_norm": 2.4211769104003906,
+      "learning_rate": 1.2095520700861948e-05,
+      "loss": 1.986,
+      "step": 39910
+    },
+    {
+      "epoch": 0.9401347086807027,
+      "grad_norm": 1.885127305984497,
+      "learning_rate": 1.2048419763553295e-05,
+      "loss": 2.0807,
+      "step": 39920
+    },
+    {
+      "epoch": 0.940370213367246,
+      "grad_norm": 2.2707126140594482,
+      "learning_rate": 1.2001318826244642e-05,
+      "loss": 2.1215,
+      "step": 39930
+    },
+    {
+      "epoch": 0.9406057180537892,
+      "grad_norm": 2.53247332572937,
+      "learning_rate": 1.195421788893599e-05,
+      "loss": 2.2294,
+      "step": 39940
+    },
+    {
+      "epoch": 0.9408412227403326,
+      "grad_norm": 2.236628532409668,
+      "learning_rate": 1.1907116951627339e-05,
+      "loss": 2.0286,
+      "step": 39950
+    },
+    {
+      "epoch": 0.9410767274268758,
+      "grad_norm": 2.0977280139923096,
+      "learning_rate": 1.1860016014318686e-05,
+      "loss": 1.9303,
+      "step": 39960
+    },
+    {
+      "epoch": 0.9413122321134191,
+      "grad_norm": 1.9017722606658936,
+      "learning_rate": 1.1812915077010033e-05,
+      "loss": 1.9777,
+      "step": 39970
+    },
+    {
+      "epoch": 0.9415477367999623,
+      "grad_norm": 2.0843467712402344,
+      "learning_rate": 1.1765814139701381e-05,
+      "loss": 2.1042,
+      "step": 39980
+    },
+    {
+      "epoch": 0.9417832414865056,
+      "grad_norm": 2.946749687194824,
+      "learning_rate": 1.1718713202392728e-05,
+      "loss": 2.0549,
+      "step": 39990
+    },
+    {
+      "epoch": 0.9420187461730488,
+      "grad_norm": 2.1347475051879883,
+      "learning_rate": 1.1671612265084075e-05,
+      "loss": 1.8955,
+      "step": 40000
+    },
+    {
+      "epoch": 0.9422542508595921,
+      "grad_norm": 3.1052968502044678,
+      "learning_rate": 1.1624511327775423e-05,
+      "loss": 2.1096,
+      "step": 40010
+    },
+    {
+      "epoch": 0.9424897555461353,
+      "grad_norm": 2.082920551300049,
+      "learning_rate": 1.157741039046677e-05,
+      "loss": 1.8824,
+      "step": 40020
+    },
+    {
+      "epoch": 0.9427252602326787,
+      "grad_norm": 2.2279791831970215,
+      "learning_rate": 1.1530309453158119e-05,
+      "loss": 2.0551,
+      "step": 40030
+    },
+    {
+      "epoch": 0.9429607649192219,
+      "grad_norm": 2.04917311668396,
+      "learning_rate": 1.1483208515849466e-05,
+      "loss": 1.9149,
+      "step": 40040
+    },
+    {
+      "epoch": 0.9431962696057652,
+      "grad_norm": 1.9186104536056519,
+      "learning_rate": 1.1436107578540814e-05,
+      "loss": 1.6671,
+      "step": 40050
+    },
+    {
+      "epoch": 0.9434317742923084,
+      "grad_norm": 2.3989248275756836,
+      "learning_rate": 1.1389006641232161e-05,
+      "loss": 2.007,
+      "step": 40060
+    },
+    {
+      "epoch": 0.9436672789788517,
+      "grad_norm": 2.3239734172821045,
+      "learning_rate": 1.1341905703923508e-05,
+      "loss": 2.064,
+      "step": 40070
+    },
+    {
+      "epoch": 0.9439027836653949,
+      "grad_norm": 1.8902684450149536,
+      "learning_rate": 1.1294804766614857e-05,
+      "loss": 2.0352,
+      "step": 40080
+    },
+    {
+      "epoch": 0.9441382883519382,
+      "grad_norm": 6.378442764282227,
+      "learning_rate": 1.1247703829306203e-05,
+      "loss": 1.9671,
+      "step": 40090
+    },
+    {
+      "epoch": 0.9443737930384815,
+      "grad_norm": 1.7690014839172363,
+      "learning_rate": 1.1200602891997552e-05,
+      "loss": 1.741,
+      "step": 40100
+    },
+    {
+      "epoch": 0.9446092977250248,
+      "grad_norm": 2.239579677581787,
+      "learning_rate": 1.1153501954688899e-05,
+      "loss": 1.8264,
+      "step": 40110
+    },
+    {
+      "epoch": 0.944844802411568,
+      "grad_norm": 2.8370306491851807,
+      "learning_rate": 1.1106401017380247e-05,
+      "loss": 1.9176,
+      "step": 40120
+    },
+    {
+      "epoch": 0.9450803070981113,
+      "grad_norm": 2.293163537979126,
+      "learning_rate": 1.1059300080071594e-05,
+      "loss": 1.9672,
+      "step": 40130
+    },
+    {
+      "epoch": 0.9453158117846545,
+      "grad_norm": 1.929100751876831,
+      "learning_rate": 1.1012199142762941e-05,
+      "loss": 1.9704,
+      "step": 40140
+    },
+    {
+      "epoch": 0.9455513164711977,
+      "grad_norm": 2.4982588291168213,
+      "learning_rate": 1.096509820545429e-05,
+      "loss": 1.925,
+      "step": 40150
+    },
+    {
+      "epoch": 0.945786821157741,
+      "grad_norm": 1.8431040048599243,
+      "learning_rate": 1.0917997268145636e-05,
+      "loss": 2.0691,
+      "step": 40160
+    },
+    {
+      "epoch": 0.9460223258442843,
+      "grad_norm": 1.8826316595077515,
+      "learning_rate": 1.0870896330836983e-05,
+      "loss": 2.2448,
+      "step": 40170
+    },
+    {
+      "epoch": 0.9462578305308276,
+      "grad_norm": 1.895750880241394,
+      "learning_rate": 1.0823795393528332e-05,
+      "loss": 2.0793,
+      "step": 40180
+    },
+    {
+      "epoch": 0.9464933352173708,
+      "grad_norm": 2.6968462467193604,
+      "learning_rate": 1.077669445621968e-05,
+      "loss": 2.0888,
+      "step": 40190
+    },
+    {
+      "epoch": 0.9467288399039141,
+      "grad_norm": 1.8779919147491455,
+      "learning_rate": 1.0729593518911027e-05,
+      "loss": 1.7711,
+      "step": 40200
+    },
+    {
+      "epoch": 0.9469643445904573,
+      "grad_norm": 2.1035165786743164,
+      "learning_rate": 1.0682492581602374e-05,
+      "loss": 1.892,
+      "step": 40210
+    },
+    {
+      "epoch": 0.9471998492770006,
+      "grad_norm": 3.0900187492370605,
+      "learning_rate": 1.0635391644293723e-05,
+      "loss": 1.9634,
+      "step": 40220
+    },
+    {
+      "epoch": 0.9474353539635438,
+      "grad_norm": 2.7870278358459473,
+      "learning_rate": 1.058829070698507e-05,
+      "loss": 2.0701,
+      "step": 40230
+    },
+    {
+      "epoch": 0.9476708586500872,
+      "grad_norm": 2.2766451835632324,
+      "learning_rate": 1.0541189769676416e-05,
+      "loss": 1.9357,
+      "step": 40240
+    },
+    {
+      "epoch": 0.9479063633366304,
+      "grad_norm": 1.974714756011963,
+      "learning_rate": 1.0494088832367765e-05,
+      "loss": 1.8973,
+      "step": 40250
+    },
+    {
+      "epoch": 0.9481418680231737,
+      "grad_norm": 2.290137767791748,
+      "learning_rate": 1.0446987895059113e-05,
+      "loss": 2.0768,
+      "step": 40260
+    },
+    {
+      "epoch": 0.9483773727097169,
+      "grad_norm": 1.5795409679412842,
+      "learning_rate": 1.039988695775046e-05,
+      "loss": 1.8673,
+      "step": 40270
+    },
+    {
+      "epoch": 0.9486128773962602,
+      "grad_norm": 2.04624605178833,
+      "learning_rate": 1.0352786020441807e-05,
+      "loss": 2.0825,
+      "step": 40280
+    },
+    {
+      "epoch": 0.9488483820828034,
+      "grad_norm": 2.657115936279297,
+      "learning_rate": 1.0305685083133154e-05,
+      "loss": 1.9653,
+      "step": 40290
+    },
+    {
+      "epoch": 0.9490838867693467,
+      "grad_norm": 1.960942029953003,
+      "learning_rate": 1.0258584145824503e-05,
+      "loss": 2.1169,
+      "step": 40300
+    },
+    {
+      "epoch": 0.94931939145589,
+      "grad_norm": 2.3039536476135254,
+      "learning_rate": 1.021148320851585e-05,
+      "loss": 1.9415,
+      "step": 40310
+    },
+    {
+      "epoch": 0.9495548961424333,
+      "grad_norm": 1.9402930736541748,
+      "learning_rate": 1.0164382271207196e-05,
+      "loss": 1.9241,
+      "step": 40320
+    },
+    {
+      "epoch": 0.9497904008289765,
+      "grad_norm": 2.1505539417266846,
+      "learning_rate": 1.0117281333898545e-05,
+      "loss": 1.8727,
+      "step": 40330
+    },
+    {
+      "epoch": 0.9500259055155198,
+      "grad_norm": 2.9176759719848633,
+      "learning_rate": 1.0070180396589893e-05,
+      "loss": 1.8201,
+      "step": 40340
+    },
+    {
+      "epoch": 0.950261410202063,
+      "grad_norm": 1.9368810653686523,
+      "learning_rate": 1.002307945928124e-05,
+      "loss": 1.9419,
+      "step": 40350
+    },
+    {
+      "epoch": 0.9504969148886063,
+      "grad_norm": 2.8279006481170654,
+      "learning_rate": 9.975978521972587e-06,
+      "loss": 1.9023,
+      "step": 40360
+    },
+    {
+      "epoch": 0.9507324195751495,
+      "grad_norm": 1.806648850440979,
+      "learning_rate": 9.928877584663936e-06,
+      "loss": 2.0467,
+      "step": 40370
+    },
+    {
+      "epoch": 0.9509679242616929,
+      "grad_norm": 2.361128807067871,
+      "learning_rate": 9.881776647355283e-06,
+      "loss": 2.1681,
+      "step": 40380
+    },
+    {
+      "epoch": 0.9512034289482361,
+      "grad_norm": 2.5333521366119385,
+      "learning_rate": 9.83467571004663e-06,
+      "loss": 1.8089,
+      "step": 40390
+    },
+    {
+      "epoch": 0.9514389336347794,
+      "grad_norm": 2.0021731853485107,
+      "learning_rate": 9.787574772737978e-06,
+      "loss": 2.0067,
+      "step": 40400
+    },
+    {
+      "epoch": 0.9516744383213226,
+      "grad_norm": 1.8866750001907349,
+      "learning_rate": 9.740473835429327e-06,
+      "loss": 2.0784,
+      "step": 40410
+    },
+    {
+      "epoch": 0.9519099430078659,
+      "grad_norm": 2.2933566570281982,
+      "learning_rate": 9.693372898120673e-06,
+      "loss": 1.9378,
+      "step": 40420
+    },
+    {
+      "epoch": 0.9521454476944091,
+      "grad_norm": 2.6166129112243652,
+      "learning_rate": 9.64627196081202e-06,
+      "loss": 1.978,
+      "step": 40430
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 2.130694627761841,
+      "learning_rate": 9.599171023503369e-06,
+      "loss": 1.9178,
+      "step": 40440
+    },
+    {
+      "epoch": 0.9526164570674956,
+      "grad_norm": 2.019354820251465,
+      "learning_rate": 9.552070086194716e-06,
+      "loss": 1.9989,
+      "step": 40450
+    },
+    {
+      "epoch": 0.9528519617540389,
+      "grad_norm": 2.3226945400238037,
+      "learning_rate": 9.504969148886062e-06,
+      "loss": 1.8383,
+      "step": 40460
+    },
+    {
+      "epoch": 0.9530874664405822,
+      "grad_norm": 3.102097511291504,
+      "learning_rate": 9.457868211577411e-06,
+      "loss": 2.1891,
+      "step": 40470
+    },
+    {
+      "epoch": 0.9533229711271254,
+      "grad_norm": 2.6576321125030518,
+      "learning_rate": 9.410767274268758e-06,
+      "loss": 2.1156,
+      "step": 40480
+    },
+    {
+      "epoch": 0.9535584758136687,
+      "grad_norm": 2.235243797302246,
+      "learning_rate": 9.363666336960106e-06,
+      "loss": 2.0167,
+      "step": 40490
+    },
+    {
+      "epoch": 0.9537939805002119,
+      "grad_norm": 4.053108215332031,
+      "learning_rate": 9.316565399651453e-06,
+      "loss": 1.975,
+      "step": 40500
+    },
+    {
+      "epoch": 0.9540294851867552,
+      "grad_norm": 3.1133532524108887,
+      "learning_rate": 9.269464462342802e-06,
+      "loss": 2.1291,
+      "step": 40510
+    },
+    {
+      "epoch": 0.9542649898732984,
+      "grad_norm": 1.7396987676620483,
+      "learning_rate": 9.222363525034149e-06,
+      "loss": 2.0171,
+      "step": 40520
+    },
+    {
+      "epoch": 0.9545004945598418,
+      "grad_norm": 2.360766649246216,
+      "learning_rate": 9.175262587725496e-06,
+      "loss": 1.8423,
+      "step": 40530
+    },
+    {
+      "epoch": 0.954735999246385,
+      "grad_norm": 1.9602024555206299,
+      "learning_rate": 9.128161650416844e-06,
+      "loss": 1.8434,
+      "step": 40540
+    },
+    {
+      "epoch": 0.9549715039329283,
+      "grad_norm": 1.9240727424621582,
+      "learning_rate": 9.081060713108191e-06,
+      "loss": 2.0884,
+      "step": 40550
+    },
+    {
+      "epoch": 0.9552070086194715,
+      "grad_norm": 1.481034517288208,
+      "learning_rate": 9.03395977579954e-06,
+      "loss": 2.115,
+      "step": 40560
+    },
+    {
+      "epoch": 0.9554425133060148,
+      "grad_norm": 2.2302894592285156,
+      "learning_rate": 8.986858838490886e-06,
+      "loss": 1.9161,
+      "step": 40570
+    },
+    {
+      "epoch": 0.955678017992558,
+      "grad_norm": 2.155421733856201,
+      "learning_rate": 8.939757901182235e-06,
+      "loss": 2.0161,
+      "step": 40580
+    },
+    {
+      "epoch": 0.9559135226791013,
+      "grad_norm": 2.0020880699157715,
+      "learning_rate": 8.892656963873582e-06,
+      "loss": 1.7731,
+      "step": 40590
+    },
+    {
+      "epoch": 0.9561490273656446,
+      "grad_norm": 2.630326271057129,
+      "learning_rate": 8.845556026564929e-06,
+      "loss": 2.0842,
+      "step": 40600
+    },
+    {
+      "epoch": 0.9563845320521879,
+      "grad_norm": 2.0237209796905518,
+      "learning_rate": 8.798455089256277e-06,
+      "loss": 1.8431,
+      "step": 40610
+    },
+    {
+      "epoch": 0.9566200367387311,
+      "grad_norm": 2.217899799346924,
+      "learning_rate": 8.751354151947624e-06,
+      "loss": 1.9634,
+      "step": 40620
+    },
+    {
+      "epoch": 0.9568555414252744,
+      "grad_norm": 2.581691026687622,
+      "learning_rate": 8.704253214638971e-06,
+      "loss": 2.1512,
+      "step": 40630
+    },
+    {
+      "epoch": 0.9570910461118176,
+      "grad_norm": 1.7218579053878784,
+      "learning_rate": 8.65715227733032e-06,
+      "loss": 1.9162,
+      "step": 40640
+    },
+    {
+      "epoch": 0.9573265507983609,
+      "grad_norm": 2.563023328781128,
+      "learning_rate": 8.610051340021668e-06,
+      "loss": 2.0915,
+      "step": 40650
+    },
+    {
+      "epoch": 0.9575620554849041,
+      "grad_norm": 2.261781692504883,
+      "learning_rate": 8.562950402713015e-06,
+      "loss": 2.2342,
+      "step": 40660
+    },
+    {
+      "epoch": 0.9577975601714475,
+      "grad_norm": 2.305806875228882,
+      "learning_rate": 8.515849465404362e-06,
+      "loss": 2.096,
+      "step": 40670
+    },
+    {
+      "epoch": 0.9580330648579907,
+      "grad_norm": 2.0696635246276855,
+      "learning_rate": 8.46874852809571e-06,
+      "loss": 2.0407,
+      "step": 40680
+    },
+    {
+      "epoch": 0.958268569544534,
+      "grad_norm": 2.0293238162994385,
+      "learning_rate": 8.421647590787057e-06,
+      "loss": 1.9653,
+      "step": 40690
+    },
+    {
+      "epoch": 0.9585040742310772,
+      "grad_norm": 1.9397081136703491,
+      "learning_rate": 8.374546653478404e-06,
+      "loss": 2.1373,
+      "step": 40700
+    },
+    {
+      "epoch": 0.9587395789176204,
+      "grad_norm": 2.025054931640625,
+      "learning_rate": 8.32744571616975e-06,
+      "loss": 2.0634,
+      "step": 40710
+    },
+    {
+      "epoch": 0.9589750836041637,
+      "grad_norm": 2.3190510272979736,
+      "learning_rate": 8.2803447788611e-06,
+      "loss": 1.8368,
+      "step": 40720
+    },
+    {
+      "epoch": 0.9592105882907069,
+      "grad_norm": 2.0201010704040527,
+      "learning_rate": 8.233243841552448e-06,
+      "loss": 1.9479,
+      "step": 40730
+    },
+    {
+      "epoch": 0.9594460929772503,
+      "grad_norm": 2.5316407680511475,
+      "learning_rate": 8.186142904243795e-06,
+      "loss": 2.0755,
+      "step": 40740
+    },
+    {
+      "epoch": 0.9596815976637935,
+      "grad_norm": 1.6399452686309814,
+      "learning_rate": 8.139041966935143e-06,
+      "loss": 1.9441,
+      "step": 40750
+    },
+    {
+      "epoch": 0.9599171023503368,
+      "grad_norm": 2.183414936065674,
+      "learning_rate": 8.09194102962649e-06,
+      "loss": 1.8982,
+      "step": 40760
+    },
+    {
+      "epoch": 0.96015260703688,
+      "grad_norm": 2.358241319656372,
+      "learning_rate": 8.044840092317837e-06,
+      "loss": 2.098,
+      "step": 40770
+    },
+    {
+      "epoch": 0.9603881117234233,
+      "grad_norm": 2.2123827934265137,
+      "learning_rate": 7.997739155009184e-06,
+      "loss": 1.9308,
+      "step": 40780
+    },
+    {
+      "epoch": 0.9606236164099665,
+      "grad_norm": 2.3104357719421387,
+      "learning_rate": 7.950638217700532e-06,
+      "loss": 2.118,
+      "step": 40790
+    },
+    {
+      "epoch": 0.9608591210965098,
+      "grad_norm": 2.0428285598754883,
+      "learning_rate": 7.903537280391881e-06,
+      "loss": 2.0509,
+      "step": 40800
+    },
+    {
+      "epoch": 0.961094625783053,
+      "grad_norm": 1.9097564220428467,
+      "learning_rate": 7.856436343083228e-06,
+      "loss": 1.8856,
+      "step": 40810
+    },
+    {
+      "epoch": 0.9613301304695964,
+      "grad_norm": 2.2115395069122314,
+      "learning_rate": 7.809335405774575e-06,
+      "loss": 1.8819,
+      "step": 40820
+    },
+    {
+      "epoch": 0.9615656351561396,
+      "grad_norm": 1.5304468870162964,
+      "learning_rate": 7.762234468465923e-06,
+      "loss": 1.7923,
+      "step": 40830
+    },
+    {
+      "epoch": 0.9618011398426829,
+      "grad_norm": 2.6274614334106445,
+      "learning_rate": 7.71513353115727e-06,
+      "loss": 2.2177,
+      "step": 40840
+    },
+    {
+      "epoch": 0.9620366445292261,
+      "grad_norm": 2.414781093597412,
+      "learning_rate": 7.668032593848617e-06,
+      "loss": 2.0156,
+      "step": 40850
+    },
+    {
+      "epoch": 0.9622721492157694,
+      "grad_norm": 2.1377460956573486,
+      "learning_rate": 7.620931656539966e-06,
+      "loss": 2.076,
+      "step": 40860
+    },
+    {
+      "epoch": 0.9625076539023126,
+      "grad_norm": 1.8874634504318237,
+      "learning_rate": 7.573830719231313e-06,
+      "loss": 2.1652,
+      "step": 40870
+    },
+    {
+      "epoch": 0.962743158588856,
+      "grad_norm": 2.2391018867492676,
+      "learning_rate": 7.52672978192266e-06,
+      "loss": 1.9276,
+      "step": 40880
+    },
+    {
+      "epoch": 0.9629786632753992,
+      "grad_norm": 2.3317453861236572,
+      "learning_rate": 7.479628844614008e-06,
+      "loss": 2.1197,
+      "step": 40890
+    },
+    {
+      "epoch": 0.9632141679619425,
+      "grad_norm": 2.019249677658081,
+      "learning_rate": 7.432527907305356e-06,
+      "loss": 2.2113,
+      "step": 40900
+    },
+    {
+      "epoch": 0.9634496726484857,
+      "grad_norm": 2.417649984359741,
+      "learning_rate": 7.385426969996703e-06,
+      "loss": 1.7919,
+      "step": 40910
+    },
+    {
+      "epoch": 0.963685177335029,
+      "grad_norm": 1.9419840574264526,
+      "learning_rate": 7.33832603268805e-06,
+      "loss": 1.8799,
+      "step": 40920
+    },
+    {
+      "epoch": 0.9639206820215722,
+      "grad_norm": 2.6572020053863525,
+      "learning_rate": 7.291225095379399e-06,
+      "loss": 2.0273,
+      "step": 40930
+    },
+    {
+      "epoch": 0.9641561867081155,
+      "grad_norm": 1.8981895446777344,
+      "learning_rate": 7.244124158070746e-06,
+      "loss": 2.0674,
+      "step": 40940
+    },
+    {
+      "epoch": 0.9643916913946587,
+      "grad_norm": 2.1581413745880127,
+      "learning_rate": 7.197023220762093e-06,
+      "loss": 1.9784,
+      "step": 40950
+    },
+    {
+      "epoch": 0.9646271960812021,
+      "grad_norm": 1.9728518724441528,
+      "learning_rate": 7.149922283453441e-06,
+      "loss": 1.9786,
+      "step": 40960
+    },
+    {
+      "epoch": 0.9648627007677453,
+      "grad_norm": 1.9783124923706055,
+      "learning_rate": 7.102821346144789e-06,
+      "loss": 1.9053,
+      "step": 40970
+    },
+    {
+      "epoch": 0.9650982054542886,
+      "grad_norm": 1.9872087240219116,
+      "learning_rate": 7.055720408836136e-06,
+      "loss": 1.867,
+      "step": 40980
+    },
+    {
+      "epoch": 0.9653337101408318,
+      "grad_norm": 2.02413010597229,
+      "learning_rate": 7.008619471527483e-06,
+      "loss": 2.0919,
+      "step": 40990
+    },
+    {
+      "epoch": 0.965569214827375,
+      "grad_norm": 2.385622978210449,
+      "learning_rate": 6.961518534218832e-06,
+      "loss": 1.9556,
+      "step": 41000
+    },
+    {
+      "epoch": 0.9658047195139183,
+      "grad_norm": 1.7820403575897217,
+      "learning_rate": 6.914417596910179e-06,
+      "loss": 1.8152,
+      "step": 41010
+    },
+    {
+      "epoch": 0.9660402242004615,
+      "grad_norm": 2.060683250427246,
+      "learning_rate": 6.867316659601526e-06,
+      "loss": 2.093,
+      "step": 41020
+    },
+    {
+      "epoch": 0.9662757288870049,
+      "grad_norm": 4.170253753662109,
+      "learning_rate": 6.820215722292873e-06,
+      "loss": 2.0762,
+      "step": 41030
+    },
+    {
+      "epoch": 0.9665112335735481,
+      "grad_norm": 2.2294230461120605,
+      "learning_rate": 6.773114784984222e-06,
+      "loss": 1.8819,
+      "step": 41040
+    },
+    {
+      "epoch": 0.9667467382600914,
+      "grad_norm": 1.990818977355957,
+      "learning_rate": 6.726013847675569e-06,
+      "loss": 2.1329,
+      "step": 41050
+    },
+    {
+      "epoch": 0.9669822429466346,
+      "grad_norm": 1.956518530845642,
+      "learning_rate": 6.678912910366916e-06,
+      "loss": 2.1007,
+      "step": 41060
+    },
+    {
+      "epoch": 0.9672177476331779,
+      "grad_norm": 2.288674831390381,
+      "learning_rate": 6.631811973058265e-06,
+      "loss": 2.0032,
+      "step": 41070
+    },
+    {
+      "epoch": 0.9674532523197211,
+      "grad_norm": 2.927011728286743,
+      "learning_rate": 6.584711035749612e-06,
+      "loss": 2.0179,
+      "step": 41080
+    },
+    {
+      "epoch": 0.9676887570062644,
+      "grad_norm": 2.0494003295898438,
+      "learning_rate": 6.537610098440959e-06,
+      "loss": 2.1701,
+      "step": 41090
+    },
+    {
+      "epoch": 0.9679242616928077,
+      "grad_norm": 1.8264861106872559,
+      "learning_rate": 6.490509161132306e-06,
+      "loss": 2.0592,
+      "step": 41100
+    },
+    {
+      "epoch": 0.968159766379351,
+      "grad_norm": 2.0611650943756104,
+      "learning_rate": 6.443408223823655e-06,
+      "loss": 2.1585,
+      "step": 41110
+    },
+    {
+      "epoch": 0.9683952710658942,
+      "grad_norm": 2.264493703842163,
+      "learning_rate": 6.396307286515002e-06,
+      "loss": 1.9042,
+      "step": 41120
+    },
+    {
+      "epoch": 0.9686307757524375,
+      "grad_norm": 2.1851072311401367,
+      "learning_rate": 6.349206349206349e-06,
+      "loss": 1.8886,
+      "step": 41130
+    },
+    {
+      "epoch": 0.9688662804389807,
+      "grad_norm": 2.544363021850586,
+      "learning_rate": 6.302105411897698e-06,
+      "loss": 2.0393,
+      "step": 41140
+    },
+    {
+      "epoch": 0.969101785125524,
+      "grad_norm": 2.145167827606201,
+      "learning_rate": 6.255004474589045e-06,
+      "loss": 2.1154,
+      "step": 41150
+    },
+    {
+      "epoch": 0.9693372898120672,
+      "grad_norm": 2.0689308643341064,
+      "learning_rate": 6.207903537280392e-06,
+      "loss": 2.2185,
+      "step": 41160
+    },
+    {
+      "epoch": 0.9695727944986106,
+      "grad_norm": 1.8527134656906128,
+      "learning_rate": 6.16080259997174e-06,
+      "loss": 1.9672,
+      "step": 41170
+    },
+    {
+      "epoch": 0.9698082991851538,
+      "grad_norm": 1.9993964433670044,
+      "learning_rate": 6.113701662663087e-06,
+      "loss": 2.0707,
+      "step": 41180
+    },
+    {
+      "epoch": 0.9700438038716971,
+      "grad_norm": 2.341604709625244,
+      "learning_rate": 6.066600725354435e-06,
+      "loss": 1.7601,
+      "step": 41190
+    },
+    {
+      "epoch": 0.9702793085582403,
+      "grad_norm": 2.557854413986206,
+      "learning_rate": 6.019499788045782e-06,
+      "loss": 2.0317,
+      "step": 41200
+    },
+    {
+      "epoch": 0.9705148132447836,
+      "grad_norm": 2.4617807865142822,
+      "learning_rate": 5.97239885073713e-06,
+      "loss": 2.0245,
+      "step": 41210
+    },
+    {
+      "epoch": 0.9707503179313268,
+      "grad_norm": 2.309467077255249,
+      "learning_rate": 5.925297913428478e-06,
+      "loss": 2.0089,
+      "step": 41220
+    },
+    {
+      "epoch": 0.9709858226178701,
+      "grad_norm": 2.5535097122192383,
+      "learning_rate": 5.878196976119825e-06,
+      "loss": 2.198,
+      "step": 41230
+    },
+    {
+      "epoch": 0.9712213273044134,
+      "grad_norm": 2.2563560009002686,
+      "learning_rate": 5.831096038811173e-06,
+      "loss": 2.2526,
+      "step": 41240
+    },
+    {
+      "epoch": 0.9714568319909567,
+      "grad_norm": 2.1398732662200928,
+      "learning_rate": 5.78399510150252e-06,
+      "loss": 1.8543,
+      "step": 41250
+    },
+    {
+      "epoch": 0.9716923366774999,
+      "grad_norm": 1.8591779470443726,
+      "learning_rate": 5.736894164193868e-06,
+      "loss": 1.8086,
+      "step": 41260
+    },
+    {
+      "epoch": 0.9719278413640432,
+      "grad_norm": 1.8352787494659424,
+      "learning_rate": 5.6897932268852146e-06,
+      "loss": 1.9678,
+      "step": 41270
+    },
+    {
+      "epoch": 0.9721633460505864,
+      "grad_norm": 2.3946144580841064,
+      "learning_rate": 5.642692289576563e-06,
+      "loss": 2.0764,
+      "step": 41280
+    },
+    {
+      "epoch": 0.9723988507371296,
+      "grad_norm": 1.8993937969207764,
+      "learning_rate": 5.595591352267911e-06,
+      "loss": 2.0317,
+      "step": 41290
+    },
+    {
+      "epoch": 0.9726343554236729,
+      "grad_norm": 1.8595757484436035,
+      "learning_rate": 5.548490414959258e-06,
+      "loss": 2.11,
+      "step": 41300
+    },
+    {
+      "epoch": 0.9728698601102161,
+      "grad_norm": 2.409909725189209,
+      "learning_rate": 5.501389477650605e-06,
+      "loss": 2.1661,
+      "step": 41310
+    },
+    {
+      "epoch": 0.9731053647967595,
+      "grad_norm": 2.140245199203491,
+      "learning_rate": 5.454288540341953e-06,
+      "loss": 1.9142,
+      "step": 41320
+    },
+    {
+      "epoch": 0.9733408694833027,
+      "grad_norm": 1.9340143203735352,
+      "learning_rate": 5.407187603033301e-06,
+      "loss": 2.1344,
+      "step": 41330
+    },
+    {
+      "epoch": 0.973576374169846,
+      "grad_norm": 2.362156867980957,
+      "learning_rate": 5.364796759455514e-06,
+      "loss": 1.6951,
+      "step": 41340
+    },
+    {
+      "epoch": 0.9738118788563892,
+      "grad_norm": 1.840448260307312,
+      "learning_rate": 5.317695822146861e-06,
+      "loss": 1.9706,
+      "step": 41350
+    },
+    {
+      "epoch": 0.9740473835429325,
+      "grad_norm": 1.9519342184066772,
+      "learning_rate": 5.270594884838208e-06,
+      "loss": 2.0591,
+      "step": 41360
+    },
+    {
+      "epoch": 0.9742828882294757,
+      "grad_norm": 1.9666930437088013,
+      "learning_rate": 5.223493947529557e-06,
+      "loss": 2.0745,
+      "step": 41370
+    },
+    {
+      "epoch": 0.974518392916019,
+      "grad_norm": 2.266784429550171,
+      "learning_rate": 5.176393010220904e-06,
+      "loss": 2.0185,
+      "step": 41380
+    },
+    {
+      "epoch": 0.9747538976025623,
+      "grad_norm": 2.675137758255005,
+      "learning_rate": 5.129292072912251e-06,
+      "loss": 1.9687,
+      "step": 41390
+    },
+    {
+      "epoch": 0.9749894022891056,
+      "grad_norm": 2.1197988986968994,
+      "learning_rate": 5.082191135603598e-06,
+      "loss": 1.9073,
+      "step": 41400
+    },
+    {
+      "epoch": 0.9752249069756488,
+      "grad_norm": 2.488123655319214,
+      "learning_rate": 5.035090198294947e-06,
+      "loss": 1.9533,
+      "step": 41410
+    },
+    {
+      "epoch": 0.9754604116621921,
+      "grad_norm": 2.570176362991333,
+      "learning_rate": 4.9879892609862936e-06,
+      "loss": 2.0849,
+      "step": 41420
+    },
+    {
+      "epoch": 0.9756959163487353,
+      "grad_norm": 2.56309175491333,
+      "learning_rate": 4.940888323677641e-06,
+      "loss": 2.1509,
+      "step": 41430
+    },
+    {
+      "epoch": 0.9759314210352786,
+      "grad_norm": 1.848645806312561,
+      "learning_rate": 4.893787386368989e-06,
+      "loss": 2.0104,
+      "step": 41440
+    },
+    {
+      "epoch": 0.9761669257218218,
+      "grad_norm": 1.9632664918899536,
+      "learning_rate": 4.846686449060337e-06,
+      "loss": 1.8198,
+      "step": 41450
+    },
+    {
+      "epoch": 0.9764024304083652,
+      "grad_norm": 2.0473501682281494,
+      "learning_rate": 4.799585511751684e-06,
+      "loss": 2.0671,
+      "step": 41460
+    },
+    {
+      "epoch": 0.9766379350949084,
+      "grad_norm": 2.3164265155792236,
+      "learning_rate": 4.752484574443031e-06,
+      "loss": 2.0143,
+      "step": 41470
+    },
+    {
+      "epoch": 0.9768734397814517,
+      "grad_norm": 2.0020151138305664,
+      "learning_rate": 4.705383637134379e-06,
+      "loss": 1.9446,
+      "step": 41480
+    },
+    {
+      "epoch": 0.9771089444679949,
+      "grad_norm": 2.8933792114257812,
+      "learning_rate": 4.658282699825727e-06,
+      "loss": 2.1086,
+      "step": 41490
+    },
+    {
+      "epoch": 0.9773444491545382,
+      "grad_norm": 1.7804278135299683,
+      "learning_rate": 4.611181762517074e-06,
+      "loss": 1.8856,
+      "step": 41500
+    },
+    {
+      "epoch": 0.9775799538410814,
+      "grad_norm": 1.9400500059127808,
+      "learning_rate": 4.564080825208422e-06,
+      "loss": 1.9144,
+      "step": 41510
+    },
+    {
+      "epoch": 0.9778154585276247,
+      "grad_norm": 2.1494195461273193,
+      "learning_rate": 4.51697988789977e-06,
+      "loss": 1.944,
+      "step": 41520
+    },
+    {
+      "epoch": 0.978050963214168,
+      "grad_norm": 1.9199514389038086,
+      "learning_rate": 4.4698789505911175e-06,
+      "loss": 1.8664,
+      "step": 41530
+    },
+    {
+      "epoch": 0.9782864679007113,
+      "grad_norm": 2.4373841285705566,
+      "learning_rate": 4.422778013282464e-06,
+      "loss": 1.9979,
+      "step": 41540
+    },
+    {
+      "epoch": 0.9785219725872545,
+      "grad_norm": 1.8263568878173828,
+      "learning_rate": 4.375677075973812e-06,
+      "loss": 2.0598,
+      "step": 41550
+    },
+    {
+      "epoch": 0.9787574772737977,
+      "grad_norm": 2.0048863887786865,
+      "learning_rate": 4.32857613866516e-06,
+      "loss": 1.8226,
+      "step": 41560
+    },
+    {
+      "epoch": 0.978992981960341,
+      "grad_norm": 2.0698230266571045,
+      "learning_rate": 4.281475201356507e-06,
+      "loss": 1.8908,
+      "step": 41570
+    },
+    {
+      "epoch": 0.9792284866468842,
+      "grad_norm": 2.3763415813446045,
+      "learning_rate": 4.234374264047855e-06,
+      "loss": 2.1119,
+      "step": 41580
+    },
+    {
+      "epoch": 0.9794639913334275,
+      "grad_norm": 2.1369435787200928,
+      "learning_rate": 4.187273326739202e-06,
+      "loss": 1.7217,
+      "step": 41590
+    },
+    {
+      "epoch": 0.9796994960199708,
+      "grad_norm": 2.2070326805114746,
+      "learning_rate": 4.14017238943055e-06,
+      "loss": 1.8161,
+      "step": 41600
+    },
+    {
+      "epoch": 0.9799350007065141,
+      "grad_norm": 2.9065499305725098,
+      "learning_rate": 4.093071452121897e-06,
+      "loss": 2.0499,
+      "step": 41610
+    },
+    {
+      "epoch": 0.9801705053930573,
+      "grad_norm": 2.06013560295105,
+      "learning_rate": 4.045970514813245e-06,
+      "loss": 1.8116,
+      "step": 41620
+    },
+    {
+      "epoch": 0.9804060100796006,
+      "grad_norm": 1.7519750595092773,
+      "learning_rate": 3.998869577504592e-06,
+      "loss": 1.9417,
+      "step": 41630
+    },
+    {
+      "epoch": 0.9806415147661438,
+      "grad_norm": 2.1897356510162354,
+      "learning_rate": 3.9517686401959405e-06,
+      "loss": 1.9161,
+      "step": 41640
+    },
+    {
+      "epoch": 0.9808770194526871,
+      "grad_norm": 2.206636428833008,
+      "learning_rate": 3.904667702887287e-06,
+      "loss": 1.8641,
+      "step": 41650
+    },
+    {
+      "epoch": 0.9811125241392303,
+      "grad_norm": 2.4233133792877197,
+      "learning_rate": 3.857566765578635e-06,
+      "loss": 1.8659,
+      "step": 41660
+    },
+    {
+      "epoch": 0.9813480288257737,
+      "grad_norm": 2.3047232627868652,
+      "learning_rate": 3.810465828269983e-06,
+      "loss": 1.9812,
+      "step": 41670
+    },
+    {
+      "epoch": 0.9815835335123169,
+      "grad_norm": 2.138484001159668,
+      "learning_rate": 3.76336489096133e-06,
+      "loss": 1.9185,
+      "step": 41680
+    },
+    {
+      "epoch": 0.9818190381988602,
+      "grad_norm": 2.0543298721313477,
+      "learning_rate": 3.716263953652678e-06,
+      "loss": 1.943,
+      "step": 41690
+    },
+    {
+      "epoch": 0.9820545428854034,
+      "grad_norm": 2.8174407482147217,
+      "learning_rate": 3.669163016344025e-06,
+      "loss": 2.2036,
+      "step": 41700
+    },
+    {
+      "epoch": 0.9822900475719467,
+      "grad_norm": 1.9986740350723267,
+      "learning_rate": 3.622062079035373e-06,
+      "loss": 1.9317,
+      "step": 41710
+    },
+    {
+      "epoch": 0.9825255522584899,
+      "grad_norm": 2.6993019580841064,
+      "learning_rate": 3.5749611417267204e-06,
+      "loss": 2.1434,
+      "step": 41720
+    },
+    {
+      "epoch": 0.9827610569450332,
+      "grad_norm": 2.0281319618225098,
+      "learning_rate": 3.527860204418068e-06,
+      "loss": 1.7839,
+      "step": 41730
+    },
+    {
+      "epoch": 0.9829965616315764,
+      "grad_norm": 2.107598304748535,
+      "learning_rate": 3.480759267109416e-06,
+      "loss": 2.0329,
+      "step": 41740
+    },
+    {
+      "epoch": 0.9832320663181198,
+      "grad_norm": 2.4774956703186035,
+      "learning_rate": 3.433658329800763e-06,
+      "loss": 2.0114,
+      "step": 41750
+    },
+    {
+      "epoch": 0.983467571004663,
+      "grad_norm": 2.693948268890381,
+      "learning_rate": 3.386557392492111e-06,
+      "loss": 2.0334,
+      "step": 41760
+    },
+    {
+      "epoch": 0.9837030756912063,
+      "grad_norm": 2.2076406478881836,
+      "learning_rate": 3.339456455183458e-06,
+      "loss": 2.1783,
+      "step": 41770
+    },
+    {
+      "epoch": 0.9839385803777495,
+      "grad_norm": 2.1578211784362793,
+      "learning_rate": 3.292355517874806e-06,
+      "loss": 1.9784,
+      "step": 41780
+    },
+    {
+      "epoch": 0.9841740850642928,
+      "grad_norm": 2.496828079223633,
+      "learning_rate": 3.245254580566153e-06,
+      "loss": 1.8877,
+      "step": 41790
+    },
+    {
+      "epoch": 0.984409589750836,
+      "grad_norm": 2.085418701171875,
+      "learning_rate": 3.198153643257501e-06,
+      "loss": 1.8715,
+      "step": 41800
+    },
+    {
+      "epoch": 0.9846450944373794,
+      "grad_norm": 2.5021634101867676,
+      "learning_rate": 3.151052705948849e-06,
+      "loss": 1.8963,
+      "step": 41810
+    },
+    {
+      "epoch": 0.9848805991239226,
+      "grad_norm": 2.096789836883545,
+      "learning_rate": 3.103951768640196e-06,
+      "loss": 1.9181,
+      "step": 41820
+    },
+    {
+      "epoch": 0.9851161038104659,
+      "grad_norm": 2.0791969299316406,
+      "learning_rate": 3.0568508313315435e-06,
+      "loss": 1.8519,
+      "step": 41830
+    },
+    {
+      "epoch": 0.9853516084970091,
+      "grad_norm": 2.2031409740448,
+      "learning_rate": 3.009749894022891e-06,
+      "loss": 2.0038,
+      "step": 41840
+    },
+    {
+      "epoch": 0.9855871131835523,
+      "grad_norm": 1.7402150630950928,
+      "learning_rate": 2.962648956714239e-06,
+      "loss": 1.968,
+      "step": 41850
+    },
+    {
+      "epoch": 0.9858226178700956,
+      "grad_norm": 1.8533766269683838,
+      "learning_rate": 2.9155480194055866e-06,
+      "loss": 2.0505,
+      "step": 41860
+    },
+    {
+      "epoch": 0.9860581225566388,
+      "grad_norm": 4.589505672454834,
+      "learning_rate": 2.868447082096934e-06,
+      "loss": 2.0808,
+      "step": 41870
+    },
+    {
+      "epoch": 0.9862936272431821,
+      "grad_norm": 1.68135404586792,
+      "learning_rate": 2.8213461447882816e-06,
+      "loss": 2.0021,
+      "step": 41880
+    },
+    {
+      "epoch": 0.9865291319297254,
+      "grad_norm": 2.261399745941162,
+      "learning_rate": 2.774245207479629e-06,
+      "loss": 1.9306,
+      "step": 41890
+    },
+    {
+      "epoch": 0.9867646366162687,
+      "grad_norm": 1.9954190254211426,
+      "learning_rate": 2.7271442701709765e-06,
+      "loss": 2.1601,
+      "step": 41900
+    },
+    {
+      "epoch": 0.9870001413028119,
+      "grad_norm": 2.2311384677886963,
+      "learning_rate": 2.680043332862324e-06,
+      "loss": 1.9097,
+      "step": 41910
+    },
+    {
+      "epoch": 0.9872356459893552,
+      "grad_norm": 1.8550012111663818,
+      "learning_rate": 2.632942395553672e-06,
+      "loss": 2.1214,
+      "step": 41920
+    },
+    {
+      "epoch": 0.9874711506758984,
+      "grad_norm": 2.2542874813079834,
+      "learning_rate": 2.5858414582450192e-06,
+      "loss": 1.8132,
+      "step": 41930
+    },
+    {
+      "epoch": 0.9877066553624417,
+      "grad_norm": 2.300264596939087,
+      "learning_rate": 2.538740520936367e-06,
+      "loss": 2.0594,
+      "step": 41940
+    },
+    {
+      "epoch": 0.9879421600489849,
+      "grad_norm": 2.213343620300293,
+      "learning_rate": 2.4916395836277142e-06,
+      "loss": 2.2302,
+      "step": 41950
+    },
+    {
+      "epoch": 0.9881776647355283,
+      "grad_norm": 2.0097548961639404,
+      "learning_rate": 2.444538646319062e-06,
+      "loss": 2.0405,
+      "step": 41960
+    },
+    {
+      "epoch": 0.9884131694220715,
+      "grad_norm": 2.0618903636932373,
+      "learning_rate": 2.397437709010409e-06,
+      "loss": 1.9674,
+      "step": 41970
+    },
+    {
+      "epoch": 0.9886486741086148,
+      "grad_norm": 2.110823631286621,
+      "learning_rate": 2.350336771701757e-06,
+      "loss": 2.25,
+      "step": 41980
+    },
+    {
+      "epoch": 0.988884178795158,
+      "grad_norm": 2.1684558391571045,
+      "learning_rate": 2.303235834393104e-06,
+      "loss": 1.8892,
+      "step": 41990
+    },
+    {
+      "epoch": 0.9891196834817013,
+      "grad_norm": 1.705733060836792,
+      "learning_rate": 2.2561348970844523e-06,
+      "loss": 1.9763,
+      "step": 42000
+    },
+    {
+      "epoch": 0.9893551881682445,
+      "grad_norm": 1.6287431716918945,
+      "learning_rate": 2.2090339597757996e-06,
+      "loss": 1.9039,
+      "step": 42010
+    },
+    {
+      "epoch": 0.9895906928547878,
+      "grad_norm": 2.455014944076538,
+      "learning_rate": 2.1619330224671473e-06,
+      "loss": 2.0905,
+      "step": 42020
+    },
+    {
+      "epoch": 0.989826197541331,
+      "grad_norm": 1.862353801727295,
+      "learning_rate": 2.1148320851584946e-06,
+      "loss": 1.9751,
+      "step": 42030
+    },
+    {
+      "epoch": 0.9900617022278744,
+      "grad_norm": 2.33832049369812,
+      "learning_rate": 2.0677311478498423e-06,
+      "loss": 1.888,
+      "step": 42040
+    },
+    {
+      "epoch": 0.9902972069144176,
+      "grad_norm": 2.076282262802124,
+      "learning_rate": 2.02063021054119e-06,
+      "loss": 1.9962,
+      "step": 42050
+    },
+    {
+      "epoch": 0.9905327116009609,
+      "grad_norm": 2.276674509048462,
+      "learning_rate": 1.9735292732325373e-06,
+      "loss": 1.9948,
+      "step": 42060
+    },
+    {
+      "epoch": 0.9907682162875041,
+      "grad_norm": 2.114805221557617,
+      "learning_rate": 1.926428335923885e-06,
+      "loss": 2.2751,
+      "step": 42070
+    },
+    {
+      "epoch": 0.9910037209740474,
+      "grad_norm": 1.974159598350525,
+      "learning_rate": 1.8793273986152327e-06,
+      "loss": 2.0404,
+      "step": 42080
+    },
+    {
+      "epoch": 0.9912392256605906,
+      "grad_norm": 2.0039196014404297,
+      "learning_rate": 1.8322264613065802e-06,
+      "loss": 2.0517,
+      "step": 42090
+    },
+    {
+      "epoch": 0.991474730347134,
+      "grad_norm": 1.9755150079727173,
+      "learning_rate": 1.7851255239979276e-06,
+      "loss": 2.1042,
+      "step": 42100
+    },
+    {
+      "epoch": 0.9917102350336772,
+      "grad_norm": 2.443861722946167,
+      "learning_rate": 1.7380245866892751e-06,
+      "loss": 2.0485,
+      "step": 42110
+    },
+    {
+      "epoch": 0.9919457397202205,
+      "grad_norm": 2.066046953201294,
+      "learning_rate": 1.6909236493806226e-06,
+      "loss": 1.917,
+      "step": 42120
+    },
+    {
+      "epoch": 0.9921812444067637,
+      "grad_norm": 2.3617396354675293,
+      "learning_rate": 1.6438227120719701e-06,
+      "loss": 1.9634,
+      "step": 42130
+    },
+    {
+      "epoch": 0.9924167490933069,
+      "grad_norm": 2.2212107181549072,
+      "learning_rate": 1.5967217747633178e-06,
+      "loss": 2.0678,
+      "step": 42140
+    },
+    {
+      "epoch": 0.9926522537798502,
+      "grad_norm": 1.842693567276001,
+      "learning_rate": 1.5496208374546653e-06,
+      "loss": 2.0056,
+      "step": 42150
+    },
+    {
+      "epoch": 0.9928877584663934,
+      "grad_norm": 2.0039405822753906,
+      "learning_rate": 1.502519900146013e-06,
+      "loss": 2.053,
+      "step": 42160
+    },
+    {
+      "epoch": 0.9931232631529368,
+      "grad_norm": 2.3391904830932617,
+      "learning_rate": 1.4554189628373605e-06,
+      "loss": 2.02,
+      "step": 42170
+    },
+    {
+      "epoch": 0.99335876783948,
+      "grad_norm": 2.3647069931030273,
+      "learning_rate": 1.4083180255287082e-06,
+      "loss": 2.0623,
+      "step": 42180
+    },
+    {
+      "epoch": 0.9935942725260233,
+      "grad_norm": 2.896048069000244,
+      "learning_rate": 1.3612170882200557e-06,
+      "loss": 2.1104,
+      "step": 42190
+    },
+    {
+      "epoch": 0.9938297772125665,
+      "grad_norm": 1.7982397079467773,
+      "learning_rate": 1.3141161509114032e-06,
+      "loss": 1.8518,
+      "step": 42200
+    },
+    {
+      "epoch": 0.9940652818991098,
+      "grad_norm": 2.1359305381774902,
+      "learning_rate": 1.2670152136027509e-06,
+      "loss": 1.9961,
+      "step": 42210
+    },
+    {
+      "epoch": 0.994300786585653,
+      "grad_norm": 1.8116016387939453,
+      "learning_rate": 1.2199142762940984e-06,
+      "loss": 2.1378,
+      "step": 42220
+    },
+    {
+      "epoch": 0.9945362912721963,
+      "grad_norm": 2.0384104251861572,
+      "learning_rate": 1.1728133389854459e-06,
+      "loss": 2.0067,
+      "step": 42230
+    },
+    {
+      "epoch": 0.9947717959587395,
+      "grad_norm": 1.8383530378341675,
+      "learning_rate": 1.1257124016767934e-06,
+      "loss": 2.0092,
+      "step": 42240
+    },
+    {
+      "epoch": 0.9950073006452829,
+      "grad_norm": 2.583355665206909,
+      "learning_rate": 1.078611464368141e-06,
+      "loss": 1.9981,
+      "step": 42250
+    },
+    {
+      "epoch": 0.9952428053318261,
+      "grad_norm": 1.9402409791946411,
+      "learning_rate": 1.0315105270594886e-06,
+      "loss": 1.9412,
+      "step": 42260
+    },
+    {
+      "epoch": 0.9954783100183694,
+      "grad_norm": 2.124499797821045,
+      "learning_rate": 9.84409589750836e-07,
+      "loss": 2.0968,
+      "step": 42270
+    },
+    {
+      "epoch": 0.9957138147049126,
+      "grad_norm": 1.9460561275482178,
+      "learning_rate": 9.373086524421838e-07,
+      "loss": 1.9956,
+      "step": 42280
+    },
+    {
+      "epoch": 0.9959493193914559,
+      "grad_norm": 2.0415124893188477,
+      "learning_rate": 8.902077151335312e-07,
+      "loss": 1.9489,
+      "step": 42290
+    },
+    {
+      "epoch": 0.9961848240779991,
+      "grad_norm": 2.739859104156494,
+      "learning_rate": 8.431067778248787e-07,
+      "loss": 2.151,
+      "step": 42300
+    },
+    {
+      "epoch": 0.9964203287645425,
+      "grad_norm": 2.373000144958496,
+      "learning_rate": 7.960058405162262e-07,
+      "loss": 1.7111,
+      "step": 42310
+    },
+    {
+      "epoch": 0.9966558334510857,
+      "grad_norm": 2.3256282806396484,
+      "learning_rate": 7.489049032075738e-07,
+      "loss": 1.9377,
+      "step": 42320
+    },
+    {
+      "epoch": 0.996891338137629,
+      "grad_norm": 2.3008763790130615,
+      "learning_rate": 7.018039658989214e-07,
+      "loss": 1.9568,
+      "step": 42330
+    },
+    {
+      "epoch": 0.9971268428241722,
+      "grad_norm": 2.1482903957366943,
+      "learning_rate": 6.547030285902689e-07,
+      "loss": 1.9677,
+      "step": 42340
+    },
+    {
+      "epoch": 0.9973623475107155,
+      "grad_norm": 2.5959582328796387,
+      "learning_rate": 6.076020912816165e-07,
+      "loss": 2.0476,
+      "step": 42350
+    },
+    {
+      "epoch": 0.9975978521972587,
+      "grad_norm": 1.8489488363265991,
+      "learning_rate": 5.605011539729641e-07,
+      "loss": 1.8465,
+      "step": 42360
+    },
+    {
+      "epoch": 0.997833356883802,
+      "grad_norm": 2.9984638690948486,
+      "learning_rate": 5.134002166643117e-07,
+      "loss": 2.171,
+      "step": 42370
+    },
+    {
+      "epoch": 0.9980688615703452,
+      "grad_norm": 1.8435646295547485,
+      "learning_rate": 4.6629927935565915e-07,
+      "loss": 1.93,
+      "step": 42380
+    },
+    {
+      "epoch": 0.9983043662568886,
+      "grad_norm": 2.5411810874938965,
+      "learning_rate": 4.1919834204700674e-07,
+      "loss": 1.9748,
+      "step": 42390
+    },
+    {
+      "epoch": 0.9985398709434318,
+      "grad_norm": 2.412177085876465,
+      "learning_rate": 3.7209740473835434e-07,
+      "loss": 1.8883,
+      "step": 42400
+    },
+    {
+      "epoch": 0.998775375629975,
+      "grad_norm": 1.8273178339004517,
+      "learning_rate": 3.249964674297019e-07,
+      "loss": 1.9645,
+      "step": 42410
+    },
+    {
+      "epoch": 0.9990108803165183,
+      "grad_norm": 2.5287861824035645,
+      "learning_rate": 2.7789553012104943e-07,
+      "loss": 1.88,
+      "step": 42420
+    },
+    {
+      "epoch": 0.9992463850030615,
+      "grad_norm": 2.011352300643921,
+      "learning_rate": 2.3079459281239697e-07,
+      "loss": 2.0222,
+      "step": 42430
+    },
+    {
+      "epoch": 0.9994818896896048,
+      "grad_norm": 2.3615148067474365,
+      "learning_rate": 1.8369365550374452e-07,
+      "loss": 1.9643,
+      "step": 42440
+    },
+    {
+      "epoch": 0.999717394376148,
+      "grad_norm": 2.0360193252563477,
+      "learning_rate": 1.365927181950921e-07,
+      "loss": 1.8166,
+      "step": 42450
+    },
+    {
+      "epoch": 0.9999528990626914,
+      "grad_norm": 2.6184275150299072,
+      "learning_rate": 8.949178088643965e-08,
+      "loss": 1.9765,
+      "step": 42460
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 42462,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "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": 4.784270210118451e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}