{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.09234050276794628,
  "eval_steps": 500,
  "global_step": 2035,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00045376168436337235,
      "grad_norm": 33.700575849311655,
      "learning_rate": 2.910112359550562e-06,
      "loss": 2.135559844970703,
      "step": 10
    },
    {
      "epoch": 0.0009075233687267447,
      "grad_norm": 42.61937790702005,
      "learning_rate": 3.921348314606742e-06,
      "loss": 1.2333361625671386,
      "step": 20
    },
    {
      "epoch": 0.001361285053090117,
      "grad_norm": 59.78398601810045,
      "learning_rate": 4.932584269662922e-06,
      "loss": 0.9913753509521485,
      "step": 30
    },
    {
      "epoch": 0.0018150467374534894,
      "grad_norm": 9.000856351949166,
      "learning_rate": 5.943820224719102e-06,
      "loss": 0.8920389175415039,
      "step": 40
    },
    {
      "epoch": 0.0022688084218168616,
      "grad_norm": 6.411342445022513,
      "learning_rate": 6.955056179775282e-06,
      "loss": 0.872259521484375,
      "step": 50
    },
    {
      "epoch": 0.002722570106180234,
      "grad_norm": 16.593767952488978,
      "learning_rate": 7.966292134831462e-06,
      "loss": 0.8293242454528809,
      "step": 60
    },
    {
      "epoch": 0.0031763317905436064,
      "grad_norm": 28.491365409414406,
      "learning_rate": 8.977528089887641e-06,
      "loss": 0.8571812629699707,
      "step": 70
    },
    {
      "epoch": 0.003630093474906979,
      "grad_norm": 70.97016809107335,
      "learning_rate": 9.988764044943822e-06,
      "loss": 0.8569000244140625,
      "step": 80
    },
    {
      "epoch": 0.004083855159270351,
      "grad_norm": 17.22979260105877,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 0.8691472053527832,
      "step": 90
    },
    {
      "epoch": 0.004537616843633723,
      "grad_norm": 7.177993416395754,
      "learning_rate": 1.2011235955056182e-05,
      "loss": 0.847593879699707,
      "step": 100
    },
    {
      "epoch": 0.004991378527997096,
      "grad_norm": 86.81590963101131,
      "learning_rate": 1.3022471910112362e-05,
      "loss": 0.7963851451873779,
      "step": 110
    },
    {
      "epoch": 0.005445140212360468,
      "grad_norm": 19.345789336537777,
      "learning_rate": 1.403370786516854e-05,
      "loss": 0.7768198013305664,
      "step": 120
    },
    {
      "epoch": 0.005898901896723841,
      "grad_norm": 31.493557815634816,
      "learning_rate": 1.504494382022472e-05,
      "loss": 0.8187647819519043,
      "step": 130
    },
    {
      "epoch": 0.006352663581087213,
      "grad_norm": 28.326685019578516,
      "learning_rate": 1.60561797752809e-05,
      "loss": 0.8060824394226074,
      "step": 140
    },
    {
      "epoch": 0.006806425265450586,
      "grad_norm": 14.750887393368782,
      "learning_rate": 1.706741573033708e-05,
      "loss": 0.8781991958618164,
      "step": 150
    },
    {
      "epoch": 0.007260186949813958,
      "grad_norm": 24.02644216435556,
      "learning_rate": 1.8078651685393256e-05,
      "loss": 0.8258254051208496,
      "step": 160
    },
    {
      "epoch": 0.00771394863417733,
      "grad_norm": 12.620566889664511,
      "learning_rate": 1.908988764044944e-05,
      "loss": 0.8197063446044922,
      "step": 170
    },
    {
      "epoch": 0.008167710318540702,
      "grad_norm": 8.507375320274459,
      "learning_rate": 2e-05,
      "loss": 0.829745864868164,
      "step": 180
    },
    {
      "epoch": 0.008621472002904074,
      "grad_norm": 8.624818937885534,
      "learning_rate": 2e-05,
      "loss": 0.8533475875854493,
      "step": 190
    },
    {
      "epoch": 0.009075233687267446,
      "grad_norm": 36.279535045904844,
      "learning_rate": 2e-05,
      "loss": 0.8423559188842773,
      "step": 200
    },
    {
      "epoch": 0.00952899537163082,
      "grad_norm": 15.3885344506724,
      "learning_rate": 2e-05,
      "loss": 0.847628116607666,
      "step": 210
    },
    {
      "epoch": 0.009982757055994192,
      "grad_norm": 6.933332738558382,
      "learning_rate": 2e-05,
      "loss": 0.8622509002685547,
      "step": 220
    },
    {
      "epoch": 0.010436518740357564,
      "grad_norm": 7.288575583436198,
      "learning_rate": 2e-05,
      "loss": 0.8516143798828125,
      "step": 230
    },
    {
      "epoch": 0.010890280424720936,
      "grad_norm": 5.432497195617162,
      "learning_rate": 2e-05,
      "loss": 0.8451622009277344,
      "step": 240
    },
    {
      "epoch": 0.01134404210908431,
      "grad_norm": 13.108630640507068,
      "learning_rate": 2e-05,
      "loss": 0.8770389556884766,
      "step": 250
    },
    {
      "epoch": 0.011797803793447682,
      "grad_norm": 13.296277537542986,
      "learning_rate": 2e-05,
      "loss": 0.8751070976257325,
      "step": 260
    },
    {
      "epoch": 0.012251565477811054,
      "grad_norm": 8.15547279712017,
      "learning_rate": 2e-05,
      "loss": 0.8359760284423828,
      "step": 270
    },
    {
      "epoch": 0.012705327162174426,
      "grad_norm": 3.8614754534197484,
      "learning_rate": 2e-05,
      "loss": 0.8397187232971192,
      "step": 280
    },
    {
      "epoch": 0.013159088846537798,
      "grad_norm": 3.928378485854922,
      "learning_rate": 2e-05,
      "loss": 0.8067515373229981,
      "step": 290
    },
    {
      "epoch": 0.013612850530901171,
      "grad_norm": 14.53890809725787,
      "learning_rate": 2e-05,
      "loss": 0.8569300651550293,
      "step": 300
    },
    {
      "epoch": 0.014066612215264543,
      "grad_norm": 2.5562493023766875,
      "learning_rate": 2e-05,
      "loss": 0.8449411392211914,
      "step": 310
    },
    {
      "epoch": 0.014520373899627915,
      "grad_norm": 2.173396163551825,
      "learning_rate": 2e-05,
      "loss": 0.8006858825683594,
      "step": 320
    },
    {
      "epoch": 0.014974135583991287,
      "grad_norm": 12.884915148556473,
      "learning_rate": 2e-05,
      "loss": 0.8169196128845215,
      "step": 330
    },
    {
      "epoch": 0.01542789726835466,
      "grad_norm": 4.133267397556853,
      "learning_rate": 2e-05,
      "loss": 0.8178797721862793,
      "step": 340
    },
    {
      "epoch": 0.015881658952718033,
      "grad_norm": 4.4839538913118,
      "learning_rate": 2e-05,
      "loss": 0.7991756439208985,
      "step": 350
    },
    {
      "epoch": 0.016335420637081405,
      "grad_norm": 5.5992645013580065,
      "learning_rate": 2e-05,
      "loss": 0.8095648765563965,
      "step": 360
    },
    {
      "epoch": 0.016789182321444777,
      "grad_norm": 2.4747669764356837,
      "learning_rate": 2e-05,
      "loss": 0.8225090026855468,
      "step": 370
    },
    {
      "epoch": 0.01724294400580815,
      "grad_norm": 2.511158852648722,
      "learning_rate": 2e-05,
      "loss": 0.7729404449462891,
      "step": 380
    },
    {
      "epoch": 0.01769670569017152,
      "grad_norm": 2.2738595224738862,
      "learning_rate": 2e-05,
      "loss": 0.7858805656433105,
      "step": 390
    },
    {
      "epoch": 0.018150467374534893,
      "grad_norm": 3.0217600696491207,
      "learning_rate": 2e-05,
      "loss": 0.8098352432250977,
      "step": 400
    },
    {
      "epoch": 0.018604229058898268,
      "grad_norm": 1.8954573295840658,
      "learning_rate": 2e-05,
      "loss": 0.7862910747528076,
      "step": 410
    },
    {
      "epoch": 0.01905799074326164,
      "grad_norm": 2.0431879466326532,
      "learning_rate": 2e-05,
      "loss": 0.7868980407714844,
      "step": 420
    },
    {
      "epoch": 0.019511752427625012,
      "grad_norm": 4.6488552474284495,
      "learning_rate": 2e-05,
      "loss": 0.7740491390228271,
      "step": 430
    },
    {
      "epoch": 0.019965514111988384,
      "grad_norm": 2.1071044234720504,
      "learning_rate": 2e-05,
      "loss": 0.7678849697113037,
      "step": 440
    },
    {
      "epoch": 0.020419275796351756,
      "grad_norm": 4.824939291325391,
      "learning_rate": 2e-05,
      "loss": 0.7809176445007324,
      "step": 450
    },
    {
      "epoch": 0.020873037480715128,
      "grad_norm": 2.3615298985348807,
      "learning_rate": 2e-05,
      "loss": 0.8083951950073243,
      "step": 460
    },
    {
      "epoch": 0.0213267991650785,
      "grad_norm": 1.3296461606887355,
      "learning_rate": 2e-05,
      "loss": 0.8065496444702148,
      "step": 470
    },
    {
      "epoch": 0.021780560849441872,
      "grad_norm": 1.8850535748077886,
      "learning_rate": 2e-05,
      "loss": 0.7924650192260743,
      "step": 480
    },
    {
      "epoch": 0.022234322533805244,
      "grad_norm": 2.8212320881753956,
      "learning_rate": 2e-05,
      "loss": 0.7533183097839355,
      "step": 490
    },
    {
      "epoch": 0.02268808421816862,
      "grad_norm": 2.1567741760252757,
      "learning_rate": 2e-05,
      "loss": 0.8023039817810058,
      "step": 500
    },
    {
      "epoch": 0.02314184590253199,
      "grad_norm": 1.913414133194118,
      "learning_rate": 2e-05,
      "loss": 0.7970404624938965,
      "step": 510
    },
    {
      "epoch": 0.023595607586895363,
      "grad_norm": 7.847343548355587,
      "learning_rate": 2e-05,
      "loss": 0.8151960372924805,
      "step": 520
    },
    {
      "epoch": 0.024049369271258735,
      "grad_norm": 1.9050240664506954,
      "learning_rate": 2e-05,
      "loss": 0.7894588470458984,
      "step": 530
    },
    {
      "epoch": 0.024503130955622107,
      "grad_norm": 1.774717889208622,
      "learning_rate": 2e-05,
      "loss": 0.7657998561859131,
      "step": 540
    },
    {
      "epoch": 0.02495689263998548,
      "grad_norm": 3.19624735773649,
      "learning_rate": 2e-05,
      "loss": 0.8172132492065429,
      "step": 550
    },
    {
      "epoch": 0.02541065432434885,
      "grad_norm": 119.89007014773632,
      "learning_rate": 2e-05,
      "loss": 0.8138757705688476,
      "step": 560
    },
    {
      "epoch": 0.025864416008712223,
      "grad_norm": 3.422241769168402,
      "learning_rate": 2e-05,
      "loss": 0.7874475479125976,
      "step": 570
    },
    {
      "epoch": 0.026318177693075595,
      "grad_norm": 30.173434363153817,
      "learning_rate": 2e-05,
      "loss": 0.8059184074401855,
      "step": 580
    },
    {
      "epoch": 0.02677193937743897,
      "grad_norm": 8.636985924843131,
      "learning_rate": 2e-05,
      "loss": 0.9973684310913086,
      "step": 590
    },
    {
      "epoch": 0.027225701061802342,
      "grad_norm": 8.465993058483805,
      "learning_rate": 2e-05,
      "loss": 0.8142690658569336,
      "step": 600
    },
    {
      "epoch": 0.027679462746165714,
      "grad_norm": 7.763176456143048,
      "learning_rate": 2e-05,
      "loss": 0.813444709777832,
      "step": 610
    },
    {
      "epoch": 0.028133224430529086,
      "grad_norm": 9.151744621497256,
      "learning_rate": 2e-05,
      "loss": 0.7639548301696777,
      "step": 620
    },
    {
      "epoch": 0.02858698611489246,
      "grad_norm": 2.0256726525570907,
      "learning_rate": 2e-05,
      "loss": 0.8043405532836914,
      "step": 630
    },
    {
      "epoch": 0.02904074779925583,
      "grad_norm": 1.98716726928425,
      "learning_rate": 2e-05,
      "loss": 0.7883163452148437,
      "step": 640
    },
    {
      "epoch": 0.029494509483619202,
      "grad_norm": 2.3604683142582186,
      "learning_rate": 2e-05,
      "loss": 0.7694703578948975,
      "step": 650
    },
    {
      "epoch": 0.029948271167982574,
      "grad_norm": 13.190838557674683,
      "learning_rate": 2e-05,
      "loss": 0.774112319946289,
      "step": 660
    },
    {
      "epoch": 0.030402032852345946,
      "grad_norm": 8.133833525009468,
      "learning_rate": 2e-05,
      "loss": 0.7486526489257812,
      "step": 670
    },
    {
      "epoch": 0.03085579453670932,
      "grad_norm": 12.970450531740925,
      "learning_rate": 2e-05,
      "loss": 0.8198841094970704,
      "step": 680
    },
    {
      "epoch": 0.03130955622107269,
      "grad_norm": 18.22753344715509,
      "learning_rate": 2e-05,
      "loss": 0.8183542251586914,
      "step": 690
    },
    {
      "epoch": 0.031763317905436066,
      "grad_norm": 5.280610592205637,
      "learning_rate": 2e-05,
      "loss": 0.8278017044067383,
      "step": 700
    },
    {
      "epoch": 0.032217079589799434,
      "grad_norm": 4.20756471706492,
      "learning_rate": 2e-05,
      "loss": 0.8102832794189453,
      "step": 710
    },
    {
      "epoch": 0.03267084127416281,
      "grad_norm": 42.72438625516298,
      "learning_rate": 2e-05,
      "loss": 0.8131640434265137,
      "step": 720
    },
    {
      "epoch": 0.033124602958526185,
      "grad_norm": 3.55536319530756,
      "learning_rate": 2e-05,
      "loss": 0.793606185913086,
      "step": 730
    },
    {
      "epoch": 0.033578364642889554,
      "grad_norm": 12.959771922355795,
      "learning_rate": 2e-05,
      "loss": 0.7831435203552246,
      "step": 740
    },
    {
      "epoch": 0.03403212632725293,
      "grad_norm": 6.469259997463987,
      "learning_rate": 2e-05,
      "loss": 0.7753517150878906,
      "step": 750
    },
    {
      "epoch": 0.0344858880116163,
      "grad_norm": 34.477603328158615,
      "learning_rate": 2e-05,
      "loss": 0.7536171913146973,
      "step": 760
    },
    {
      "epoch": 0.03493964969597967,
      "grad_norm": 2.328509918402065,
      "learning_rate": 2e-05,
      "loss": 0.7809432029724122,
      "step": 770
    },
    {
      "epoch": 0.03539341138034304,
      "grad_norm": 4.288352937474369,
      "learning_rate": 2e-05,
      "loss": 0.7952579498291016,
      "step": 780
    },
    {
      "epoch": 0.03584717306470642,
      "grad_norm": 7.639274296461497,
      "learning_rate": 2e-05,
      "loss": 0.8167254447937011,
      "step": 790
    },
    {
      "epoch": 0.036300934749069785,
      "grad_norm": 2.5903744589165703,
      "learning_rate": 2e-05,
      "loss": 0.7940209865570068,
      "step": 800
    },
    {
      "epoch": 0.03675469643343316,
      "grad_norm": 1.714046167478395,
      "learning_rate": 2e-05,
      "loss": 0.7802738189697266,
      "step": 810
    },
    {
      "epoch": 0.037208458117796536,
      "grad_norm": 2.356191303105163,
      "learning_rate": 2e-05,
      "loss": 0.7603988170623779,
      "step": 820
    },
    {
      "epoch": 0.037662219802159905,
      "grad_norm": 2.6324392358450206,
      "learning_rate": 2e-05,
      "loss": 0.7551675319671631,
      "step": 830
    },
    {
      "epoch": 0.03811598148652328,
      "grad_norm": 5.1149592386678036,
      "learning_rate": 2e-05,
      "loss": 0.7713553428649902,
      "step": 840
    },
    {
      "epoch": 0.03856974317088665,
      "grad_norm": 4.250591266250503,
      "learning_rate": 2e-05,
      "loss": 0.7647957801818848,
      "step": 850
    },
    {
      "epoch": 0.039023504855250024,
      "grad_norm": 1.484157740941218,
      "learning_rate": 2e-05,
      "loss": 0.7875026226043701,
      "step": 860
    },
    {
      "epoch": 0.03947726653961339,
      "grad_norm": 2.4537172207690814,
      "learning_rate": 2e-05,
      "loss": 0.7982926368713379,
      "step": 870
    },
    {
      "epoch": 0.03993102822397677,
      "grad_norm": 1.885761099206628,
      "learning_rate": 2e-05,
      "loss": 0.7709912776947021,
      "step": 880
    },
    {
      "epoch": 0.04038478990834014,
      "grad_norm": 3.1726728603243513,
      "learning_rate": 2e-05,
      "loss": 0.7637313842773438,
      "step": 890
    },
    {
      "epoch": 0.04083855159270351,
      "grad_norm": 1.734967497093215,
      "learning_rate": 2e-05,
      "loss": 0.78067626953125,
      "step": 900
    },
    {
      "epoch": 0.04129231327706689,
      "grad_norm": 3.298790323372369,
      "learning_rate": 2e-05,
      "loss": 0.7815995693206788,
      "step": 910
    },
    {
      "epoch": 0.041746074961430256,
      "grad_norm": 1.3197318794126558,
      "learning_rate": 2e-05,
      "loss": 0.7714223861694336,
      "step": 920
    },
    {
      "epoch": 0.04219983664579363,
      "grad_norm": 2.179324182696462,
      "learning_rate": 2e-05,
      "loss": 0.752001142501831,
      "step": 930
    },
    {
      "epoch": 0.042653598330157,
      "grad_norm": 1.7760762886617165,
      "learning_rate": 2e-05,
      "loss": 0.7619918346405029,
      "step": 940
    },
    {
      "epoch": 0.043107360014520375,
      "grad_norm": 5.1168406185966235,
      "learning_rate": 2e-05,
      "loss": 0.7750636577606201,
      "step": 950
    },
    {
      "epoch": 0.043561121698883744,
      "grad_norm": 1.8080346226599653,
      "learning_rate": 2e-05,
      "loss": 0.7890705108642578,
      "step": 960
    },
    {
      "epoch": 0.04401488338324712,
      "grad_norm": 1.5578518869917433,
      "learning_rate": 2e-05,
      "loss": 0.7533071041107178,
      "step": 970
    },
    {
      "epoch": 0.04446864506761049,
      "grad_norm": 2.0331640286970094,
      "learning_rate": 2e-05,
      "loss": 0.7617393493652344,
      "step": 980
    },
    {
      "epoch": 0.04492240675197386,
      "grad_norm": 1.2015052654892246,
      "learning_rate": 2e-05,
      "loss": 0.7665273189544678,
      "step": 990
    },
    {
      "epoch": 0.04537616843633724,
      "grad_norm": 1.628143392881603,
      "learning_rate": 2e-05,
      "loss": 0.770779037475586,
      "step": 1000
    },
    {
      "epoch": 0.04582993012070061,
      "grad_norm": 1.2039882827957658,
      "learning_rate": 2e-05,
      "loss": 0.7887059211730957,
      "step": 1010
    },
    {
      "epoch": 0.04628369180506398,
      "grad_norm": 1.5827671273859805,
      "learning_rate": 2e-05,
      "loss": 0.7503167629241944,
      "step": 1020
    },
    {
      "epoch": 0.04673745348942735,
      "grad_norm": 3.868510887527835,
      "learning_rate": 2e-05,
      "loss": 0.7796614646911622,
      "step": 1030
    },
    {
      "epoch": 0.047191215173790727,
      "grad_norm": 2.9140174925180067,
      "learning_rate": 2e-05,
      "loss": 0.8053103446960449,
      "step": 1040
    },
    {
      "epoch": 0.047644976858154095,
      "grad_norm": 7.892541947564107,
      "learning_rate": 2e-05,
      "loss": 0.8123416900634766,
      "step": 1050
    },
    {
      "epoch": 0.04809873854251747,
      "grad_norm": 1.8993885865680917,
      "learning_rate": 2e-05,
      "loss": 0.7850803375244141,
      "step": 1060
    },
    {
      "epoch": 0.04855250022688084,
      "grad_norm": 2.1292285710118657,
      "learning_rate": 2e-05,
      "loss": 0.8199052810668945,
      "step": 1070
    },
    {
      "epoch": 0.049006261911244214,
      "grad_norm": 1.4765373586924606,
      "learning_rate": 2e-05,
      "loss": 0.7806931972503662,
      "step": 1080
    },
    {
      "epoch": 0.04946002359560759,
      "grad_norm": 1.516040890469186,
      "learning_rate": 2e-05,
      "loss": 0.7693392753601074,
      "step": 1090
    },
    {
      "epoch": 0.04991378527997096,
      "grad_norm": 2.0831299604400417,
      "learning_rate": 2e-05,
      "loss": 0.7863246917724609,
      "step": 1100
    },
    {
      "epoch": 0.050367546964334334,
      "grad_norm": 2.195827470530011,
      "learning_rate": 2e-05,
      "loss": 0.7650362968444824,
      "step": 1110
    },
    {
      "epoch": 0.0508213086486977,
      "grad_norm": 3.203619514755763,
      "learning_rate": 2e-05,
      "loss": 0.7856436729431152,
      "step": 1120
    },
    {
      "epoch": 0.05127507033306108,
      "grad_norm": 3.564960215102263,
      "learning_rate": 2e-05,
      "loss": 0.7714583396911621,
      "step": 1130
    },
    {
      "epoch": 0.051728832017424446,
      "grad_norm": 2.602524659508837,
      "learning_rate": 2e-05,
      "loss": 0.7595517635345459,
      "step": 1140
    },
    {
      "epoch": 0.05218259370178782,
      "grad_norm": 3.0105932895678977,
      "learning_rate": 2e-05,
      "loss": 0.7478479385375977,
      "step": 1150
    },
    {
      "epoch": 0.05263635538615119,
      "grad_norm": 1.9912075171919272,
      "learning_rate": 2e-05,
      "loss": 0.7397075176239014,
      "step": 1160
    },
    {
      "epoch": 0.053090117070514566,
      "grad_norm": 3.0032576981429444,
      "learning_rate": 2e-05,
      "loss": 0.7824847221374511,
      "step": 1170
    },
    {
      "epoch": 0.05354387875487794,
      "grad_norm": 1.7741279587695722,
      "learning_rate": 2e-05,
      "loss": 0.7740449428558349,
      "step": 1180
    },
    {
      "epoch": 0.05399764043924131,
      "grad_norm": 1.9161385528021593,
      "learning_rate": 2e-05,
      "loss": 0.7688920021057128,
      "step": 1190
    },
    {
      "epoch": 0.054451402123604685,
      "grad_norm": 1.6376294125787574,
      "learning_rate": 2e-05,
      "loss": 0.7435236930847168,
      "step": 1200
    },
    {
      "epoch": 0.05490516380796805,
      "grad_norm": 2.655987836587518,
      "learning_rate": 2e-05,
      "loss": 0.7560546875,
      "step": 1210
    },
    {
      "epoch": 0.05535892549233143,
      "grad_norm": 1.5087730766928222,
      "learning_rate": 2e-05,
      "loss": 0.7511416435241699,
      "step": 1220
    },
    {
      "epoch": 0.0558126871766948,
      "grad_norm": 2.9588632994615547,
      "learning_rate": 2e-05,
      "loss": 0.7484208583831787,
      "step": 1230
    },
    {
      "epoch": 0.05626644886105817,
      "grad_norm": 13.660542975087127,
      "learning_rate": 2e-05,
      "loss": 0.7785249710083008,
      "step": 1240
    },
    {
      "epoch": 0.05672021054542154,
      "grad_norm": 3.189563629774851,
      "learning_rate": 2e-05,
      "loss": 0.7544436454772949,
      "step": 1250
    },
    {
      "epoch": 0.05717397222978492,
      "grad_norm": 23.612106104880457,
      "learning_rate": 2e-05,
      "loss": 0.7925362586975098,
      "step": 1260
    },
    {
      "epoch": 0.05762773391414829,
      "grad_norm": 3.377156755225946,
      "learning_rate": 2e-05,
      "loss": 0.7938989162445068,
      "step": 1270
    },
    {
      "epoch": 0.05808149559851166,
      "grad_norm": 1.6453871136214189,
      "learning_rate": 2e-05,
      "loss": 0.7648170471191407,
      "step": 1280
    },
    {
      "epoch": 0.058535257282875036,
      "grad_norm": 1.7098215646357884,
      "learning_rate": 2e-05,
      "loss": 0.7756949424743652,
      "step": 1290
    },
    {
      "epoch": 0.058989018967238405,
      "grad_norm": 3.6636095771340043,
      "learning_rate": 2e-05,
      "loss": 0.8291525840759277,
      "step": 1300
    },
    {
      "epoch": 0.05944278065160178,
      "grad_norm": 2.177902708692178,
      "learning_rate": 2e-05,
      "loss": 0.7537973403930665,
      "step": 1310
    },
    {
      "epoch": 0.05989654233596515,
      "grad_norm": 2.1786512930926945,
      "learning_rate": 2e-05,
      "loss": 0.7512475490570069,
      "step": 1320
    },
    {
      "epoch": 0.060350304020328524,
      "grad_norm": 2.3889515043153953,
      "learning_rate": 2e-05,
      "loss": 0.7963077068328858,
      "step": 1330
    },
    {
      "epoch": 0.06080406570469189,
      "grad_norm": 6.049406950820628,
      "learning_rate": 2e-05,
      "loss": 0.7398816585540772,
      "step": 1340
    },
    {
      "epoch": 0.06125782738905527,
      "grad_norm": 1.6910884112612445,
      "learning_rate": 2e-05,
      "loss": 0.7709364891052246,
      "step": 1350
    },
    {
      "epoch": 0.06171158907341864,
      "grad_norm": 2.355501745859088,
      "learning_rate": 2e-05,
      "loss": 0.7542487621307373,
      "step": 1360
    },
    {
      "epoch": 0.06216535075778201,
      "grad_norm": 2.1903603191775782,
      "learning_rate": 2e-05,
      "loss": 0.8148787498474122,
      "step": 1370
    },
    {
      "epoch": 0.06261911244214538,
      "grad_norm": 4.583080645477956,
      "learning_rate": 2e-05,
      "loss": 0.7330617427825927,
      "step": 1380
    },
    {
      "epoch": 0.06307287412650876,
      "grad_norm": 1.5890129050702244,
      "learning_rate": 2e-05,
      "loss": 0.7422567367553711,
      "step": 1390
    },
    {
      "epoch": 0.06352663581087213,
      "grad_norm": 1.163231824259212,
      "learning_rate": 2e-05,
      "loss": 0.7530087947845459,
      "step": 1400
    },
    {
      "epoch": 0.0639803974952355,
      "grad_norm": 1.9648397918825078,
      "learning_rate": 2e-05,
      "loss": 0.7386720180511475,
      "step": 1410
    },
    {
      "epoch": 0.06443415917959887,
      "grad_norm": 1.1154033888302555,
      "learning_rate": 2e-05,
      "loss": 0.7690919399261474,
      "step": 1420
    },
    {
      "epoch": 0.06488792086396225,
      "grad_norm": 2.817335534651707,
      "learning_rate": 2e-05,
      "loss": 0.7828889846801758,
      "step": 1430
    },
    {
      "epoch": 0.06534168254832562,
      "grad_norm": 1.5652197356391113,
      "learning_rate": 2e-05,
      "loss": 0.7418618202209473,
      "step": 1440
    },
    {
      "epoch": 0.06579544423268899,
      "grad_norm": 1.4323144167320692,
      "learning_rate": 2e-05,
      "loss": 0.7751660823822022,
      "step": 1450
    },
    {
      "epoch": 0.06624920591705237,
      "grad_norm": 2.0558277244752623,
      "learning_rate": 2e-05,
      "loss": 0.7712644577026367,
      "step": 1460
    },
    {
      "epoch": 0.06670296760141574,
      "grad_norm": 19.677486710967138,
      "learning_rate": 2e-05,
      "loss": 0.7851467132568359,
      "step": 1470
    },
    {
      "epoch": 0.06715672928577911,
      "grad_norm": 25.384982473704017,
      "learning_rate": 2e-05,
      "loss": 0.7457910537719726,
      "step": 1480
    },
    {
      "epoch": 0.06761049097014248,
      "grad_norm": 1.1405705553998693,
      "learning_rate": 2e-05,
      "loss": 0.7223224639892578,
      "step": 1490
    },
    {
      "epoch": 0.06806425265450586,
      "grad_norm": 1.5986852504273663,
      "learning_rate": 2e-05,
      "loss": 0.7437433242797852,
      "step": 1500
    },
    {
      "epoch": 0.06851801433886923,
      "grad_norm": 1.473364803589886,
      "learning_rate": 2e-05,
      "loss": 0.7854445934295654,
      "step": 1510
    },
    {
      "epoch": 0.0689717760232326,
      "grad_norm": 10.816907441073242,
      "learning_rate": 2e-05,
      "loss": 0.7771403312683105,
      "step": 1520
    },
    {
      "epoch": 0.06942553770759598,
      "grad_norm": 2.3871999487587425,
      "learning_rate": 2e-05,
      "loss": 0.7488942623138428,
      "step": 1530
    },
    {
      "epoch": 0.06987929939195935,
      "grad_norm": 1.29078615018717,
      "learning_rate": 2e-05,
      "loss": 0.7702004432678222,
      "step": 1540
    },
    {
      "epoch": 0.07033306107632271,
      "grad_norm": 1.2427006882362452,
      "learning_rate": 2e-05,
      "loss": 0.7769896507263183,
      "step": 1550
    },
    {
      "epoch": 0.07078682276068608,
      "grad_norm": 1.8501499468782703,
      "learning_rate": 2e-05,
      "loss": 0.7520946025848388,
      "step": 1560
    },
    {
      "epoch": 0.07124058444504947,
      "grad_norm": 2.600210573565917,
      "learning_rate": 2e-05,
      "loss": 0.7633906841278076,
      "step": 1570
    },
    {
      "epoch": 0.07169434612941283,
      "grad_norm": 2.522289120850679,
      "learning_rate": 2e-05,
      "loss": 0.7302196502685547,
      "step": 1580
    },
    {
      "epoch": 0.0721481078137762,
      "grad_norm": 1.3515878124531273,
      "learning_rate": 2e-05,
      "loss": 0.7600698471069336,
      "step": 1590
    },
    {
      "epoch": 0.07260186949813957,
      "grad_norm": 1.140634184395665,
      "learning_rate": 2e-05,
      "loss": 0.7558602333068848,
      "step": 1600
    },
    {
      "epoch": 0.07305563118250295,
      "grad_norm": 1.3170969672293371,
      "learning_rate": 2e-05,
      "loss": 0.7792963027954102,
      "step": 1610
    },
    {
      "epoch": 0.07350939286686632,
      "grad_norm": 12.865316269450114,
      "learning_rate": 2e-05,
      "loss": 0.7376986980438233,
      "step": 1620
    },
    {
      "epoch": 0.07396315455122969,
      "grad_norm": 1.2524502323072844,
      "learning_rate": 2e-05,
      "loss": 0.7631961822509765,
      "step": 1630
    },
    {
      "epoch": 0.07441691623559307,
      "grad_norm": 1.1643893423939573,
      "learning_rate": 2e-05,
      "loss": 0.7555351257324219,
      "step": 1640
    },
    {
      "epoch": 0.07487067791995644,
      "grad_norm": 3.1366425147206507,
      "learning_rate": 2e-05,
      "loss": 0.7184365749359131,
      "step": 1650
    },
    {
      "epoch": 0.07532443960431981,
      "grad_norm": 1.6802684630791902,
      "learning_rate": 2e-05,
      "loss": 0.716264295578003,
      "step": 1660
    },
    {
      "epoch": 0.07577820128868318,
      "grad_norm": 1.4738450371157292,
      "learning_rate": 2e-05,
      "loss": 0.7608924865722656,
      "step": 1670
    },
    {
      "epoch": 0.07623196297304656,
      "grad_norm": 3.932860483703015,
      "learning_rate": 2e-05,
      "loss": 0.763948917388916,
      "step": 1680
    },
    {
      "epoch": 0.07668572465740993,
      "grad_norm": 1.6528466518745035,
      "learning_rate": 2e-05,
      "loss": 0.7618756771087647,
      "step": 1690
    },
    {
      "epoch": 0.0771394863417733,
      "grad_norm": 0.9245158103125077,
      "learning_rate": 2e-05,
      "loss": 0.7731586456298828,
      "step": 1700
    },
    {
      "epoch": 0.07759324802613668,
      "grad_norm": 1.3297115851518835,
      "learning_rate": 2e-05,
      "loss": 0.7360218048095704,
      "step": 1710
    },
    {
      "epoch": 0.07804700971050005,
      "grad_norm": 1.1975864742002036,
      "learning_rate": 2e-05,
      "loss": 0.7769705772399902,
      "step": 1720
    },
    {
      "epoch": 0.07850077139486342,
      "grad_norm": 1.2410115385284919,
      "learning_rate": 2e-05,
      "loss": 0.7371129035949707,
      "step": 1730
    },
    {
      "epoch": 0.07895453307922679,
      "grad_norm": 1.7317801478129846,
      "learning_rate": 2e-05,
      "loss": 0.7660929679870605,
      "step": 1740
    },
    {
      "epoch": 0.07940829476359017,
      "grad_norm": 0.9051550955007733,
      "learning_rate": 2e-05,
      "loss": 0.7519168376922607,
      "step": 1750
    },
    {
      "epoch": 0.07986205644795354,
      "grad_norm": 1.6723802945669413,
      "learning_rate": 2e-05,
      "loss": 0.7212412834167481,
      "step": 1760
    },
    {
      "epoch": 0.0803158181323169,
      "grad_norm": 0.9762152447317929,
      "learning_rate": 2e-05,
      "loss": 0.7753897666931152,
      "step": 1770
    },
    {
      "epoch": 0.08076957981668027,
      "grad_norm": 1.2473869795723174,
      "learning_rate": 2e-05,
      "loss": 0.7250448226928711,
      "step": 1780
    },
    {
      "epoch": 0.08122334150104366,
      "grad_norm": 3.1833001654859108,
      "learning_rate": 2e-05,
      "loss": 0.7206393241882324,
      "step": 1790
    },
    {
      "epoch": 0.08167710318540702,
      "grad_norm": 1.595724497195245,
      "learning_rate": 2e-05,
      "loss": 0.7571566581726075,
      "step": 1800
    },
    {
      "epoch": 0.08213086486977039,
      "grad_norm": 1.4572318958737844,
      "learning_rate": 2e-05,
      "loss": 0.7492036342620849,
      "step": 1810
    },
    {
      "epoch": 0.08258462655413377,
      "grad_norm": 2.696706475039584,
      "learning_rate": 2e-05,
      "loss": 0.757764196395874,
      "step": 1820
    },
    {
      "epoch": 0.08303838823849714,
      "grad_norm": 9.705539395813444,
      "learning_rate": 2e-05,
      "loss": 0.7491719245910644,
      "step": 1830
    },
    {
      "epoch": 0.08349214992286051,
      "grad_norm": 2.445157502577676,
      "learning_rate": 2e-05,
      "loss": 0.7391232967376709,
      "step": 1840
    },
    {
      "epoch": 0.08394591160722388,
      "grad_norm": 2.2218722227847745,
      "learning_rate": 2e-05,
      "loss": 0.7663413047790527,
      "step": 1850
    },
    {
      "epoch": 0.08439967329158726,
      "grad_norm": 6.437451692035588,
      "learning_rate": 2e-05,
      "loss": 0.7212537288665771,
      "step": 1860
    },
    {
      "epoch": 0.08485343497595063,
      "grad_norm": 14.459355718334681,
      "learning_rate": 2e-05,
      "loss": 0.7453588008880615,
      "step": 1870
    },
    {
      "epoch": 0.085307196660314,
      "grad_norm": 2.274387051434555,
      "learning_rate": 2e-05,
      "loss": 0.7336775779724121,
      "step": 1880
    },
    {
      "epoch": 0.08576095834467738,
      "grad_norm": 2.0275002345768827,
      "learning_rate": 2e-05,
      "loss": 0.8232375144958496,
      "step": 1890
    },
    {
      "epoch": 0.08621472002904075,
      "grad_norm": 3.4079005741269714,
      "learning_rate": 2e-05,
      "loss": 0.7306999206542969,
      "step": 1900
    },
    {
      "epoch": 0.08666848171340412,
      "grad_norm": 2.888480518708026,
      "learning_rate": 2e-05,
      "loss": 0.7341453552246093,
      "step": 1910
    },
    {
      "epoch": 0.08712224339776749,
      "grad_norm": 2.691347164562704,
      "learning_rate": 2e-05,
      "loss": 0.7440048217773437,
      "step": 1920
    },
    {
      "epoch": 0.08757600508213087,
      "grad_norm": 2.1392692969810208,
      "learning_rate": 2e-05,
      "loss": 0.7886202812194825,
      "step": 1930
    },
    {
      "epoch": 0.08802976676649424,
      "grad_norm": 3.1332404791177275,
      "learning_rate": 2e-05,
      "loss": 0.7390870571136474,
      "step": 1940
    },
    {
      "epoch": 0.08848352845085761,
      "grad_norm": 2.2360847842037592,
      "learning_rate": 2e-05,
      "loss": 0.7575728416442871,
      "step": 1950
    },
    {
      "epoch": 0.08893729013522098,
      "grad_norm": 4.969135746166459,
      "learning_rate": 2e-05,
      "loss": 0.734520435333252,
      "step": 1960
    },
    {
      "epoch": 0.08939105181958436,
      "grad_norm": 3.046212464054818,
      "learning_rate": 2e-05,
      "loss": 0.7530054092407227,
      "step": 1970
    },
    {
      "epoch": 0.08984481350394773,
      "grad_norm": 4.929848622369691,
      "learning_rate": 2e-05,
      "loss": 0.7513354301452637,
      "step": 1980
    },
    {
      "epoch": 0.0902985751883111,
      "grad_norm": 3.8207364141673184,
      "learning_rate": 2e-05,
      "loss": 0.7366059780120849,
      "step": 1990
    },
    {
      "epoch": 0.09075233687267448,
      "grad_norm": 3.445571619835256,
      "learning_rate": 2e-05,
      "loss": 0.7999159812927246,
      "step": 2000
    },
    {
      "epoch": 0.09120609855703785,
      "grad_norm": 2.210897089598567,
      "learning_rate": 2e-05,
      "loss": 0.7329919338226318,
      "step": 2010
    },
    {
      "epoch": 0.09165986024140121,
      "grad_norm": 1.3151585193661326,
      "learning_rate": 2e-05,
      "loss": 0.7405821800231933,
      "step": 2020
    },
    {
      "epoch": 0.09211362192576458,
      "grad_norm": 3.175239796553618,
      "learning_rate": 2e-05,
      "loss": 0.7426975250244141,
      "step": 2030
    }
  ],
  "logging_steps": 10,
  "max_steps": 5935,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 2035,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 963613493821440.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}