rv2-gemma12b-g2_5_plan_newcode / final /trainer_state.json
rqadri's picture
Upload folder using huggingface_hub
7f1f627 verified
Raw History Blame Contribute Delete
103 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.26930755966966147,
"eval_steps": 500,
"global_step": 5935,
"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
},
{
"epoch": 0.09256738361012797,
"grad_norm": 1.5460986491899282,
"learning_rate": 2e-05,
"loss": 0.7336253166198731,
"step": 2040
},
{
"epoch": 0.09302114529449133,
"grad_norm": 2.44617108722647,
"learning_rate": 2e-05,
"loss": 0.7323103904724121,
"step": 2050
},
{
"epoch": 0.0934749069788547,
"grad_norm": 1.2023342852694456,
"learning_rate": 2e-05,
"loss": 0.7493421554565429,
"step": 2060
},
{
"epoch": 0.09392866866321808,
"grad_norm": 1.6501372967395636,
"learning_rate": 2e-05,
"loss": 0.7366260051727295,
"step": 2070
},
{
"epoch": 0.09438243034758145,
"grad_norm": 1.5630609314797599,
"learning_rate": 2e-05,
"loss": 0.7564666748046875,
"step": 2080
},
{
"epoch": 0.09483619203194482,
"grad_norm": 1.100093524345851,
"learning_rate": 2e-05,
"loss": 0.7178175926208497,
"step": 2090
},
{
"epoch": 0.09528995371630819,
"grad_norm": 1.203687520284051,
"learning_rate": 2e-05,
"loss": 0.8053166389465332,
"step": 2100
},
{
"epoch": 0.09574371540067157,
"grad_norm": 1.755317345775532,
"learning_rate": 2e-05,
"loss": 0.7951737403869629,
"step": 2110
},
{
"epoch": 0.09619747708503494,
"grad_norm": 1.1331278156810007,
"learning_rate": 2e-05,
"loss": 0.7216485977172852,
"step": 2120
},
{
"epoch": 0.09665123876939831,
"grad_norm": 1.3863522587743262,
"learning_rate": 2e-05,
"loss": 0.7505239486694336,
"step": 2130
},
{
"epoch": 0.09710500045376168,
"grad_norm": 1.106142535654746,
"learning_rate": 2e-05,
"loss": 0.7577492713928222,
"step": 2140
},
{
"epoch": 0.09755876213812506,
"grad_norm": 1.0991599118921827,
"learning_rate": 2e-05,
"loss": 0.7133097648620605,
"step": 2150
},
{
"epoch": 0.09801252382248843,
"grad_norm": 1.2943390794331155,
"learning_rate": 2e-05,
"loss": 0.7233659267425537,
"step": 2160
},
{
"epoch": 0.0984662855068518,
"grad_norm": 1.0502136600603071,
"learning_rate": 2e-05,
"loss": 0.7315991401672364,
"step": 2170
},
{
"epoch": 0.09892004719121518,
"grad_norm": 1.4068742926964601,
"learning_rate": 2e-05,
"loss": 0.8028385162353515,
"step": 2180
},
{
"epoch": 0.09937380887557855,
"grad_norm": 2.534737268552639,
"learning_rate": 2e-05,
"loss": 0.7296046733856201,
"step": 2190
},
{
"epoch": 0.09982757055994192,
"grad_norm": 1.2151364240607376,
"learning_rate": 2e-05,
"loss": 0.7711813926696778,
"step": 2200
},
{
"epoch": 0.10028133224430529,
"grad_norm": 12.637697159869406,
"learning_rate": 2e-05,
"loss": 0.7253233909606933,
"step": 2210
},
{
"epoch": 0.10073509392866867,
"grad_norm": 4.187414972043617,
"learning_rate": 2e-05,
"loss": 0.7317253112792969,
"step": 2220
},
{
"epoch": 0.10118885561303204,
"grad_norm": 6.009696438025872,
"learning_rate": 2e-05,
"loss": 0.7461653709411621,
"step": 2230
},
{
"epoch": 0.1016426172973954,
"grad_norm": 1.5320622144398235,
"learning_rate": 2e-05,
"loss": 0.7306909561157227,
"step": 2240
},
{
"epoch": 0.10209637898175879,
"grad_norm": 1.154341731917967,
"learning_rate": 2e-05,
"loss": 0.7285379409790039,
"step": 2250
},
{
"epoch": 0.10255014066612216,
"grad_norm": 1.2061265003483534,
"learning_rate": 2e-05,
"loss": 0.7660748481750488,
"step": 2260
},
{
"epoch": 0.10300390235048552,
"grad_norm": 1.2663031896068002,
"learning_rate": 2e-05,
"loss": 0.7520371913909912,
"step": 2270
},
{
"epoch": 0.10345766403484889,
"grad_norm": 2.5817180740910617,
"learning_rate": 2e-05,
"loss": 0.7496843338012695,
"step": 2280
},
{
"epoch": 0.10391142571921227,
"grad_norm": 1.1804254908468201,
"learning_rate": 2e-05,
"loss": 0.6986440181732178,
"step": 2290
},
{
"epoch": 0.10436518740357564,
"grad_norm": 2.3525473107366137,
"learning_rate": 2e-05,
"loss": 0.7352646827697754,
"step": 2300
},
{
"epoch": 0.10481894908793901,
"grad_norm": 1.1147715673751997,
"learning_rate": 2e-05,
"loss": 0.714558219909668,
"step": 2310
},
{
"epoch": 0.10527271077230238,
"grad_norm": 2.9561178864781366,
"learning_rate": 2e-05,
"loss": 0.7822277069091796,
"step": 2320
},
{
"epoch": 0.10572647245666576,
"grad_norm": 5.743993808519714,
"learning_rate": 2e-05,
"loss": 0.7459967613220215,
"step": 2330
},
{
"epoch": 0.10618023414102913,
"grad_norm": 1.715356350906201,
"learning_rate": 2e-05,
"loss": 0.7326190948486329,
"step": 2340
},
{
"epoch": 0.1066339958253925,
"grad_norm": 7.329409059571893,
"learning_rate": 2e-05,
"loss": 0.7490138530731201,
"step": 2350
},
{
"epoch": 0.10708775750975588,
"grad_norm": 5.898819908652382,
"learning_rate": 2e-05,
"loss": 0.7567214012145996,
"step": 2360
},
{
"epoch": 0.10754151919411925,
"grad_norm": 3.324367151744097,
"learning_rate": 2e-05,
"loss": 0.7694614410400391,
"step": 2370
},
{
"epoch": 0.10799528087848262,
"grad_norm": 2.942329415798371,
"learning_rate": 2e-05,
"loss": 0.730305290222168,
"step": 2380
},
{
"epoch": 0.10844904256284599,
"grad_norm": 2.6406465368358747,
"learning_rate": 2e-05,
"loss": 0.7168290615081787,
"step": 2390
},
{
"epoch": 0.10890280424720937,
"grad_norm": 1.6013660041922653,
"learning_rate": 2e-05,
"loss": 0.7251017570495606,
"step": 2400
},
{
"epoch": 0.10935656593157274,
"grad_norm": 2.9192570486679807,
"learning_rate": 2e-05,
"loss": 0.8059085845947266,
"step": 2410
},
{
"epoch": 0.1098103276159361,
"grad_norm": 1.8675022460247312,
"learning_rate": 2e-05,
"loss": 0.7660941123962403,
"step": 2420
},
{
"epoch": 0.11026408930029949,
"grad_norm": 3.9337815260869546,
"learning_rate": 2e-05,
"loss": 0.732635498046875,
"step": 2430
},
{
"epoch": 0.11071785098466286,
"grad_norm": 1.9972292386385522,
"learning_rate": 2e-05,
"loss": 0.7119313240051269,
"step": 2440
},
{
"epoch": 0.11117161266902623,
"grad_norm": 3.488921868612139,
"learning_rate": 2e-05,
"loss": 0.7181739807128906,
"step": 2450
},
{
"epoch": 0.1116253743533896,
"grad_norm": 4.122806602217105,
"learning_rate": 2e-05,
"loss": 0.7661437034606934,
"step": 2460
},
{
"epoch": 0.11207913603775298,
"grad_norm": 1.7018225863042173,
"learning_rate": 2e-05,
"loss": 0.7263262748718262,
"step": 2470
},
{
"epoch": 0.11253289772211635,
"grad_norm": 1.9300439188671026,
"learning_rate": 2e-05,
"loss": 0.7203723907470703,
"step": 2480
},
{
"epoch": 0.11298665940647971,
"grad_norm": 5.589563878507127,
"learning_rate": 2e-05,
"loss": 0.7311852455139161,
"step": 2490
},
{
"epoch": 0.11344042109084308,
"grad_norm": 1.410438934945161,
"learning_rate": 2e-05,
"loss": 0.7646050930023194,
"step": 2500
},
{
"epoch": 0.11389418277520647,
"grad_norm": 2.8165074482881955,
"learning_rate": 2e-05,
"loss": 0.7460585594177246,
"step": 2510
},
{
"epoch": 0.11434794445956983,
"grad_norm": 1.5142120250689506,
"learning_rate": 2e-05,
"loss": 0.7251346588134766,
"step": 2520
},
{
"epoch": 0.1148017061439332,
"grad_norm": 2.840694299031808,
"learning_rate": 2e-05,
"loss": 0.730368185043335,
"step": 2530
},
{
"epoch": 0.11525546782829658,
"grad_norm": 1.2855707947922668,
"learning_rate": 2e-05,
"loss": 0.7467219352722168,
"step": 2540
},
{
"epoch": 0.11570922951265995,
"grad_norm": 11.664631653248765,
"learning_rate": 2e-05,
"loss": 0.720964002609253,
"step": 2550
},
{
"epoch": 0.11616299119702332,
"grad_norm": 1.071483909274107,
"learning_rate": 2e-05,
"loss": 0.7421702861785888,
"step": 2560
},
{
"epoch": 0.11661675288138669,
"grad_norm": 0.9494600172925889,
"learning_rate": 2e-05,
"loss": 0.7440818786621094,
"step": 2570
},
{
"epoch": 0.11707051456575007,
"grad_norm": 1.1662851842589446,
"learning_rate": 2e-05,
"loss": 0.740445613861084,
"step": 2580
},
{
"epoch": 0.11752427625011344,
"grad_norm": 1.4530177337648706,
"learning_rate": 2e-05,
"loss": 0.7615899562835693,
"step": 2590
},
{
"epoch": 0.11797803793447681,
"grad_norm": 1.1215699653059057,
"learning_rate": 2e-05,
"loss": 0.7278817653656006,
"step": 2600
},
{
"epoch": 0.11843179961884019,
"grad_norm": 3.1171434879670716,
"learning_rate": 2e-05,
"loss": 0.7337120056152344,
"step": 2610
},
{
"epoch": 0.11888556130320356,
"grad_norm": 1.5121308129364615,
"learning_rate": 2e-05,
"loss": 0.7112399101257324,
"step": 2620
},
{
"epoch": 0.11933932298756693,
"grad_norm": 1.8713224495325613,
"learning_rate": 2e-05,
"loss": 0.7495639324188232,
"step": 2630
},
{
"epoch": 0.1197930846719303,
"grad_norm": 2.021004325138582,
"learning_rate": 2e-05,
"loss": 0.7113505363464355,
"step": 2640
},
{
"epoch": 0.12024684635629368,
"grad_norm": 3.040606817980415,
"learning_rate": 2e-05,
"loss": 0.72547607421875,
"step": 2650
},
{
"epoch": 0.12070060804065705,
"grad_norm": 2.6860589524294793,
"learning_rate": 2e-05,
"loss": 0.716103458404541,
"step": 2660
},
{
"epoch": 0.12115436972502042,
"grad_norm": 2.071747558666727,
"learning_rate": 2e-05,
"loss": 0.6948127746582031,
"step": 2670
},
{
"epoch": 0.12160813140938379,
"grad_norm": 2.0311162196387964,
"learning_rate": 2e-05,
"loss": 0.7409467697143555,
"step": 2680
},
{
"epoch": 0.12206189309374717,
"grad_norm": 3.117906624141895,
"learning_rate": 2e-05,
"loss": 0.7186509609222412,
"step": 2690
},
{
"epoch": 0.12251565477811054,
"grad_norm": 1.4745245098436555,
"learning_rate": 2e-05,
"loss": 0.7544195175170898,
"step": 2700
},
{
"epoch": 0.1229694164624739,
"grad_norm": 1.3142703663159712,
"learning_rate": 2e-05,
"loss": 0.7209731578826905,
"step": 2710
},
{
"epoch": 0.12342317814683729,
"grad_norm": 1.6099723331957767,
"learning_rate": 2e-05,
"loss": 0.7600280284881592,
"step": 2720
},
{
"epoch": 0.12387693983120066,
"grad_norm": 1.7538718564922926,
"learning_rate": 2e-05,
"loss": 0.7137237548828125,
"step": 2730
},
{
"epoch": 0.12433070151556402,
"grad_norm": 2.2073523165352493,
"learning_rate": 2e-05,
"loss": 0.7364648342132568,
"step": 2740
},
{
"epoch": 0.12478446319992739,
"grad_norm": 1.588870611678532,
"learning_rate": 2e-05,
"loss": 0.7471172332763671,
"step": 2750
},
{
"epoch": 0.12523822488429076,
"grad_norm": 1.451885876651238,
"learning_rate": 2e-05,
"loss": 0.7264121055603028,
"step": 2760
},
{
"epoch": 0.12569198656865413,
"grad_norm": 1.3375977416884601,
"learning_rate": 2e-05,
"loss": 0.7405541419982911,
"step": 2770
},
{
"epoch": 0.12614574825301753,
"grad_norm": 1.4287617860448656,
"learning_rate": 2e-05,
"loss": 0.7860785484313965,
"step": 2780
},
{
"epoch": 0.1265995099373809,
"grad_norm": 1.1768557257747507,
"learning_rate": 2e-05,
"loss": 0.761849308013916,
"step": 2790
},
{
"epoch": 0.12705327162174426,
"grad_norm": 1.0074979786836618,
"learning_rate": 2e-05,
"loss": 0.7187026023864747,
"step": 2800
},
{
"epoch": 0.12750703330610763,
"grad_norm": 1.1780816826485483,
"learning_rate": 2e-05,
"loss": 0.7368823051452636,
"step": 2810
},
{
"epoch": 0.127960794990471,
"grad_norm": 1.498407817125236,
"learning_rate": 2e-05,
"loss": 0.7312613487243652,
"step": 2820
},
{
"epoch": 0.12841455667483437,
"grad_norm": 1.8319692829426788,
"learning_rate": 2e-05,
"loss": 0.7246813774108887,
"step": 2830
},
{
"epoch": 0.12886831835919774,
"grad_norm": 1.0834014899051665,
"learning_rate": 2e-05,
"loss": 0.7331777095794678,
"step": 2840
},
{
"epoch": 0.12932208004356113,
"grad_norm": 1.6849144829970544,
"learning_rate": 2e-05,
"loss": 0.7145289421081543,
"step": 2850
},
{
"epoch": 0.1297758417279245,
"grad_norm": 1.2160780024942408,
"learning_rate": 2e-05,
"loss": 0.7284758567810059,
"step": 2860
},
{
"epoch": 0.13022960341228787,
"grad_norm": 1.6383395416823836,
"learning_rate": 2e-05,
"loss": 0.7749518394470215,
"step": 2870
},
{
"epoch": 0.13068336509665124,
"grad_norm": 1.2032477916661783,
"learning_rate": 2e-05,
"loss": 0.7386165618896484,
"step": 2880
},
{
"epoch": 0.1311371267810146,
"grad_norm": 1.1053484705577017,
"learning_rate": 2e-05,
"loss": 0.7096535682678222,
"step": 2890
},
{
"epoch": 0.13159088846537798,
"grad_norm": 1.0244402834052642,
"learning_rate": 2e-05,
"loss": 0.7102260589599609,
"step": 2900
},
{
"epoch": 0.13204465014974134,
"grad_norm": 1.0734139909623037,
"learning_rate": 2e-05,
"loss": 0.7314916133880616,
"step": 2910
},
{
"epoch": 0.13249841183410474,
"grad_norm": 0.9695130146801769,
"learning_rate": 2e-05,
"loss": 0.7415484428405762,
"step": 2920
},
{
"epoch": 0.1329521735184681,
"grad_norm": 0.9693439274108523,
"learning_rate": 2e-05,
"loss": 0.712250804901123,
"step": 2930
},
{
"epoch": 0.13340593520283148,
"grad_norm": 4.41282937974211,
"learning_rate": 2e-05,
"loss": 0.7414229869842529,
"step": 2940
},
{
"epoch": 0.13385969688719485,
"grad_norm": 1.0491399353510926,
"learning_rate": 2e-05,
"loss": 0.7940215110778809,
"step": 2950
},
{
"epoch": 0.13431345857155821,
"grad_norm": 0.978631304878327,
"learning_rate": 2e-05,
"loss": 0.7373782157897949,
"step": 2960
},
{
"epoch": 0.13476722025592158,
"grad_norm": 1.0856533220063784,
"learning_rate": 2e-05,
"loss": 0.7301129817962646,
"step": 2970
},
{
"epoch": 0.13522098194028495,
"grad_norm": 1.0909698875981262,
"learning_rate": 2e-05,
"loss": 0.7138740062713623,
"step": 2980
},
{
"epoch": 0.13567474362464835,
"grad_norm": 1.323562425128533,
"learning_rate": 2e-05,
"loss": 0.7477345943450928,
"step": 2990
},
{
"epoch": 0.13612850530901172,
"grad_norm": 0.9673584767071247,
"learning_rate": 2e-05,
"loss": 0.7181426048278808,
"step": 3000
},
{
"epoch": 0.13658226699337508,
"grad_norm": 2.651566918809511,
"learning_rate": 2e-05,
"loss": 0.7557494163513183,
"step": 3010
},
{
"epoch": 0.13703602867773845,
"grad_norm": 1.8013290501367338,
"learning_rate": 2e-05,
"loss": 0.7161075115203858,
"step": 3020
},
{
"epoch": 0.13748979036210182,
"grad_norm": 1.2040252299685656,
"learning_rate": 2e-05,
"loss": 0.7047561645507813,
"step": 3030
},
{
"epoch": 0.1379435520464652,
"grad_norm": 1.0654594031315352,
"learning_rate": 2e-05,
"loss": 0.6870278358459473,
"step": 3040
},
{
"epoch": 0.13839731373082856,
"grad_norm": 0.9366966272340514,
"learning_rate": 2e-05,
"loss": 0.7171539306640625,
"step": 3050
},
{
"epoch": 0.13885107541519195,
"grad_norm": 1.181805133773245,
"learning_rate": 2e-05,
"loss": 0.7271934509277344,
"step": 3060
},
{
"epoch": 0.13930483709955532,
"grad_norm": 1.0263383968648079,
"learning_rate": 2e-05,
"loss": 0.7353593826293945,
"step": 3070
},
{
"epoch": 0.1397585987839187,
"grad_norm": 1.1418288655004183,
"learning_rate": 2e-05,
"loss": 0.6901812553405762,
"step": 3080
},
{
"epoch": 0.14021236046828206,
"grad_norm": 1.2838260319859818,
"learning_rate": 2e-05,
"loss": 0.7208109855651855,
"step": 3090
},
{
"epoch": 0.14066612215264543,
"grad_norm": 2.7593588187273155,
"learning_rate": 2e-05,
"loss": 0.7018424510955811,
"step": 3100
},
{
"epoch": 0.1411198838370088,
"grad_norm": 1.1631736464965061,
"learning_rate": 2e-05,
"loss": 0.7176068782806396,
"step": 3110
},
{
"epoch": 0.14157364552137217,
"grad_norm": 1.1076860739441265,
"learning_rate": 2e-05,
"loss": 0.6939721584320069,
"step": 3120
},
{
"epoch": 0.14202740720573553,
"grad_norm": 1.078154119908151,
"learning_rate": 2e-05,
"loss": 0.7411942958831788,
"step": 3130
},
{
"epoch": 0.14248116889009893,
"grad_norm": 0.9761645612675013,
"learning_rate": 2e-05,
"loss": 0.792600154876709,
"step": 3140
},
{
"epoch": 0.1429349305744623,
"grad_norm": 1.2170191765233065,
"learning_rate": 2e-05,
"loss": 0.7225303649902344,
"step": 3150
},
{
"epoch": 0.14338869225882567,
"grad_norm": 1.1352052491760056,
"learning_rate": 2e-05,
"loss": 0.7216647148132325,
"step": 3160
},
{
"epoch": 0.14384245394318904,
"grad_norm": 1.0045691377146662,
"learning_rate": 2e-05,
"loss": 0.7115875244140625,
"step": 3170
},
{
"epoch": 0.1442962156275524,
"grad_norm": 1.0543653950112917,
"learning_rate": 2e-05,
"loss": 0.7357659816741944,
"step": 3180
},
{
"epoch": 0.14474997731191577,
"grad_norm": 1.0504240446734914,
"learning_rate": 2e-05,
"loss": 0.684734296798706,
"step": 3190
},
{
"epoch": 0.14520373899627914,
"grad_norm": 0.8051170281000236,
"learning_rate": 2e-05,
"loss": 0.7431095123291016,
"step": 3200
},
{
"epoch": 0.14565750068064254,
"grad_norm": 1.1524866017575786,
"learning_rate": 2e-05,
"loss": 0.7744518280029297,
"step": 3210
},
{
"epoch": 0.1461112623650059,
"grad_norm": 1.0089112922281773,
"learning_rate": 2e-05,
"loss": 0.7189767360687256,
"step": 3220
},
{
"epoch": 0.14656502404936927,
"grad_norm": 1.1626904315580493,
"learning_rate": 2e-05,
"loss": 0.697879695892334,
"step": 3230
},
{
"epoch": 0.14701878573373264,
"grad_norm": 1.033357799949316,
"learning_rate": 2e-05,
"loss": 0.7418144226074219,
"step": 3240
},
{
"epoch": 0.147472547418096,
"grad_norm": 0.839481568059364,
"learning_rate": 2e-05,
"loss": 0.6834076881408692,
"step": 3250
},
{
"epoch": 0.14792630910245938,
"grad_norm": 1.0459052369304707,
"learning_rate": 2e-05,
"loss": 0.7019244194030761,
"step": 3260
},
{
"epoch": 0.14838007078682275,
"grad_norm": 1.040827164215203,
"learning_rate": 2e-05,
"loss": 0.700564956665039,
"step": 3270
},
{
"epoch": 0.14883383247118614,
"grad_norm": 1.1352996331387266,
"learning_rate": 2e-05,
"loss": 0.7144912719726563,
"step": 3280
},
{
"epoch": 0.1492875941555495,
"grad_norm": 1.7425202277535818,
"learning_rate": 2e-05,
"loss": 0.7080129146575928,
"step": 3290
},
{
"epoch": 0.14974135583991288,
"grad_norm": 2.8379547304992987,
"learning_rate": 2e-05,
"loss": 0.7066136837005615,
"step": 3300
},
{
"epoch": 0.15019511752427625,
"grad_norm": 1.104417859965814,
"learning_rate": 2e-05,
"loss": 0.7218207359313965,
"step": 3310
},
{
"epoch": 0.15064887920863962,
"grad_norm": 1.339952671768318,
"learning_rate": 2e-05,
"loss": 0.7319831848144531,
"step": 3320
},
{
"epoch": 0.151102640893003,
"grad_norm": 1.6764918067862147,
"learning_rate": 2e-05,
"loss": 0.7054910659790039,
"step": 3330
},
{
"epoch": 0.15155640257736636,
"grad_norm": 1.5638697746561578,
"learning_rate": 2e-05,
"loss": 0.7452583312988281,
"step": 3340
},
{
"epoch": 0.15201016426172975,
"grad_norm": 1.3945883708700582,
"learning_rate": 2e-05,
"loss": 0.7258811950683594,
"step": 3350
},
{
"epoch": 0.15246392594609312,
"grad_norm": 1.018126375950651,
"learning_rate": 2e-05,
"loss": 0.7109189510345459,
"step": 3360
},
{
"epoch": 0.1529176876304565,
"grad_norm": 1.1715830697545735,
"learning_rate": 2e-05,
"loss": 0.7115130424499512,
"step": 3370
},
{
"epoch": 0.15337144931481986,
"grad_norm": 1.4837142351649248,
"learning_rate": 2e-05,
"loss": 0.7141985893249512,
"step": 3380
},
{
"epoch": 0.15382521099918323,
"grad_norm": 1.3594143043972335,
"learning_rate": 2e-05,
"loss": 0.7526227474212647,
"step": 3390
},
{
"epoch": 0.1542789726835466,
"grad_norm": 1.2536102303932697,
"learning_rate": 2e-05,
"loss": 0.7404181957244873,
"step": 3400
},
{
"epoch": 0.15473273436790996,
"grad_norm": 1.3611412930630256,
"learning_rate": 2e-05,
"loss": 0.7230888366699219,
"step": 3410
},
{
"epoch": 0.15518649605227336,
"grad_norm": 1.1363549729780036,
"learning_rate": 2e-05,
"loss": 0.6954388618469238,
"step": 3420
},
{
"epoch": 0.15564025773663673,
"grad_norm": 1.1062316759863606,
"learning_rate": 2e-05,
"loss": 0.7978248596191406,
"step": 3430
},
{
"epoch": 0.1560940194210001,
"grad_norm": 1.0748709736048603,
"learning_rate": 2e-05,
"loss": 0.7171708583831787,
"step": 3440
},
{
"epoch": 0.15654778110536346,
"grad_norm": 1.1941060643552694,
"learning_rate": 2e-05,
"loss": 0.7279775142669678,
"step": 3450
},
{
"epoch": 0.15700154278972683,
"grad_norm": 0.9832361767157862,
"learning_rate": 2e-05,
"loss": 0.6856719017028808,
"step": 3460
},
{
"epoch": 0.1574553044740902,
"grad_norm": 0.9314871875656874,
"learning_rate": 2e-05,
"loss": 0.7153595924377442,
"step": 3470
},
{
"epoch": 0.15790906615845357,
"grad_norm": 0.9763506764277187,
"learning_rate": 2e-05,
"loss": 0.6920379638671875,
"step": 3480
},
{
"epoch": 0.15836282784281694,
"grad_norm": 1.5434706616412097,
"learning_rate": 2e-05,
"loss": 0.7263721466064453,
"step": 3490
},
{
"epoch": 0.15881658952718034,
"grad_norm": 1.1384979168118918,
"learning_rate": 2e-05,
"loss": 0.7131462097167969,
"step": 3500
},
{
"epoch": 0.1592703512115437,
"grad_norm": 1.3200596768314172,
"learning_rate": 2e-05,
"loss": 0.7170546531677247,
"step": 3510
},
{
"epoch": 0.15972411289590707,
"grad_norm": 1.5818006592270006,
"learning_rate": 2e-05,
"loss": 0.7287190914154053,
"step": 3520
},
{
"epoch": 0.16017787458027044,
"grad_norm": 1.2510371962585238,
"learning_rate": 2e-05,
"loss": 0.7110270500183106,
"step": 3530
},
{
"epoch": 0.1606316362646338,
"grad_norm": 0.9093620537994762,
"learning_rate": 2e-05,
"loss": 0.7264317989349365,
"step": 3540
},
{
"epoch": 0.16108539794899718,
"grad_norm": 1.4410904531435618,
"learning_rate": 2e-05,
"loss": 0.7267816066741943,
"step": 3550
},
{
"epoch": 0.16153915963336055,
"grad_norm": 1.0813912067522184,
"learning_rate": 2e-05,
"loss": 0.751043701171875,
"step": 3560
},
{
"epoch": 0.16199292131772394,
"grad_norm": 1.0522037520194414,
"learning_rate": 2e-05,
"loss": 0.7142654418945312,
"step": 3570
},
{
"epoch": 0.1624466830020873,
"grad_norm": 0.9622624873032466,
"learning_rate": 2e-05,
"loss": 0.6926068782806396,
"step": 3580
},
{
"epoch": 0.16290044468645068,
"grad_norm": 1.28463962211785,
"learning_rate": 2e-05,
"loss": 0.708470630645752,
"step": 3590
},
{
"epoch": 0.16335420637081405,
"grad_norm": 1.020953581409204,
"learning_rate": 2e-05,
"loss": 0.7161651611328125,
"step": 3600
},
{
"epoch": 0.16380796805517742,
"grad_norm": 0.9351542221824213,
"learning_rate": 2e-05,
"loss": 0.7340415000915528,
"step": 3610
},
{
"epoch": 0.16426172973954078,
"grad_norm": 0.865040770646376,
"learning_rate": 2e-05,
"loss": 0.7186861991882324,
"step": 3620
},
{
"epoch": 0.16471549142390415,
"grad_norm": 1.4568579992964554,
"learning_rate": 2e-05,
"loss": 0.7627389907836915,
"step": 3630
},
{
"epoch": 0.16516925310826755,
"grad_norm": 1.2305715937510664,
"learning_rate": 2e-05,
"loss": 0.722710371017456,
"step": 3640
},
{
"epoch": 0.16562301479263092,
"grad_norm": 0.8283750793341599,
"learning_rate": 2e-05,
"loss": 0.7440120697021484,
"step": 3650
},
{
"epoch": 0.1660767764769943,
"grad_norm": 0.9752798522028159,
"learning_rate": 2e-05,
"loss": 0.7054687023162842,
"step": 3660
},
{
"epoch": 0.16653053816135766,
"grad_norm": 5.206653791148631,
"learning_rate": 2e-05,
"loss": 0.7331564903259278,
"step": 3670
},
{
"epoch": 0.16698429984572102,
"grad_norm": 0.9751607647203324,
"learning_rate": 2e-05,
"loss": 0.7228295803070068,
"step": 3680
},
{
"epoch": 0.1674380615300844,
"grad_norm": 3.4399584110011365,
"learning_rate": 2e-05,
"loss": 0.700427770614624,
"step": 3690
},
{
"epoch": 0.16789182321444776,
"grad_norm": 1.733120495772477,
"learning_rate": 2e-05,
"loss": 0.7794057369232178,
"step": 3700
},
{
"epoch": 0.16834558489881116,
"grad_norm": 1.6264083682750483,
"learning_rate": 2e-05,
"loss": 0.7170550346374511,
"step": 3710
},
{
"epoch": 0.16879934658317453,
"grad_norm": 1.2746147227893116,
"learning_rate": 2e-05,
"loss": 0.6994328975677491,
"step": 3720
},
{
"epoch": 0.1692531082675379,
"grad_norm": 1.7975240289797993,
"learning_rate": 2e-05,
"loss": 0.7199362754821778,
"step": 3730
},
{
"epoch": 0.16970686995190126,
"grad_norm": 1.6942005697688205,
"learning_rate": 2e-05,
"loss": 0.6994699478149414,
"step": 3740
},
{
"epoch": 0.17016063163626463,
"grad_norm": 1.6546419917677124,
"learning_rate": 2e-05,
"loss": 0.7093739986419678,
"step": 3750
},
{
"epoch": 0.170614393320628,
"grad_norm": 1.546903486384874,
"learning_rate": 2e-05,
"loss": 0.7056426048278809,
"step": 3760
},
{
"epoch": 0.17106815500499137,
"grad_norm": 1.3582344043489818,
"learning_rate": 2e-05,
"loss": 0.7326198577880859,
"step": 3770
},
{
"epoch": 0.17152191668935476,
"grad_norm": 1.3183359355415163,
"learning_rate": 2e-05,
"loss": 0.7046661376953125,
"step": 3780
},
{
"epoch": 0.17197567837371813,
"grad_norm": 3.7327801389540434,
"learning_rate": 2e-05,
"loss": 0.6876080989837646,
"step": 3790
},
{
"epoch": 0.1724294400580815,
"grad_norm": 2.0357177183751114,
"learning_rate": 2e-05,
"loss": 0.6858410835266113,
"step": 3800
},
{
"epoch": 0.17288320174244487,
"grad_norm": 1.227416992036067,
"learning_rate": 2e-05,
"loss": 0.8117395401000976,
"step": 3810
},
{
"epoch": 0.17333696342680824,
"grad_norm": 1.18242287111851,
"learning_rate": 2e-05,
"loss": 0.7055291652679443,
"step": 3820
},
{
"epoch": 0.1737907251111716,
"grad_norm": 1.002486271567816,
"learning_rate": 2e-05,
"loss": 0.6894384384155273,
"step": 3830
},
{
"epoch": 0.17424448679553498,
"grad_norm": 1.1271488657415134,
"learning_rate": 2e-05,
"loss": 0.7097995758056641,
"step": 3840
},
{
"epoch": 0.17469824847989834,
"grad_norm": 1.2251097105201723,
"learning_rate": 2e-05,
"loss": 0.7608026504516602,
"step": 3850
},
{
"epoch": 0.17515201016426174,
"grad_norm": 2.104712995855096,
"learning_rate": 2e-05,
"loss": 0.7532779216766358,
"step": 3860
},
{
"epoch": 0.1756057718486251,
"grad_norm": 1.242711378221653,
"learning_rate": 2e-05,
"loss": 0.7006797790527344,
"step": 3870
},
{
"epoch": 0.17605953353298848,
"grad_norm": 1.0241935233576007,
"learning_rate": 2e-05,
"loss": 0.737086820602417,
"step": 3880
},
{
"epoch": 0.17651329521735185,
"grad_norm": 1.6359170007243453,
"learning_rate": 2e-05,
"loss": 0.6877423763275147,
"step": 3890
},
{
"epoch": 0.17696705690171521,
"grad_norm": 1.405430278807483,
"learning_rate": 2e-05,
"loss": 0.6851609230041504,
"step": 3900
},
{
"epoch": 0.17742081858607858,
"grad_norm": 6.752700891821782,
"learning_rate": 2e-05,
"loss": 0.7518776893615723,
"step": 3910
},
{
"epoch": 0.17787458027044195,
"grad_norm": 2.610510657458775,
"learning_rate": 2e-05,
"loss": 0.7406556129455566,
"step": 3920
},
{
"epoch": 0.17832834195480535,
"grad_norm": 1.0484775936839292,
"learning_rate": 2e-05,
"loss": 0.737001895904541,
"step": 3930
},
{
"epoch": 0.17878210363916872,
"grad_norm": 1.3323103500916966,
"learning_rate": 2e-05,
"loss": 0.7060594081878662,
"step": 3940
},
{
"epoch": 0.17923586532353208,
"grad_norm": 1.5130187182517638,
"learning_rate": 2e-05,
"loss": 0.7092230796813965,
"step": 3950
},
{
"epoch": 0.17968962700789545,
"grad_norm": 3.6503019384420408,
"learning_rate": 2e-05,
"loss": 0.7280755043029785,
"step": 3960
},
{
"epoch": 0.18014338869225882,
"grad_norm": 1.116592675851622,
"learning_rate": 2e-05,
"loss": 0.7035638332366944,
"step": 3970
},
{
"epoch": 0.1805971503766222,
"grad_norm": 1.142280213213852,
"learning_rate": 2e-05,
"loss": 0.7515734672546387,
"step": 3980
},
{
"epoch": 0.18105091206098556,
"grad_norm": 1.8155022158547836,
"learning_rate": 2e-05,
"loss": 0.6944926738739013,
"step": 3990
},
{
"epoch": 0.18150467374534895,
"grad_norm": 1.3297567956941603,
"learning_rate": 2e-05,
"loss": 0.7093602657318115,
"step": 4000
},
{
"epoch": 0.18195843542971232,
"grad_norm": 2.686017230236506,
"learning_rate": 2e-05,
"loss": 0.7569149017333985,
"step": 4010
},
{
"epoch": 0.1824121971140757,
"grad_norm": 1.0389215623357642,
"learning_rate": 2e-05,
"loss": 0.6734948635101319,
"step": 4020
},
{
"epoch": 0.18286595879843906,
"grad_norm": 5.216030343061308,
"learning_rate": 2e-05,
"loss": 0.7012555599212646,
"step": 4030
},
{
"epoch": 0.18331972048280243,
"grad_norm": 4.024862852173242,
"learning_rate": 2e-05,
"loss": 0.7192629814147949,
"step": 4040
},
{
"epoch": 0.1837734821671658,
"grad_norm": 1.1885454364724406,
"learning_rate": 2e-05,
"loss": 0.6805631160736084,
"step": 4050
},
{
"epoch": 0.18422724385152917,
"grad_norm": 1.0174535627554133,
"learning_rate": 2e-05,
"loss": 0.7393642425537109,
"step": 4060
},
{
"epoch": 0.18468100553589256,
"grad_norm": 1.2902170593905826,
"learning_rate": 2e-05,
"loss": 0.7127656459808349,
"step": 4070
},
{
"epoch": 0.18513476722025593,
"grad_norm": 1.2571358195560212,
"learning_rate": 2e-05,
"loss": 0.6989442348480225,
"step": 4080
},
{
"epoch": 0.1855885289046193,
"grad_norm": 1.082663616523454,
"learning_rate": 2e-05,
"loss": 0.6818115234375,
"step": 4090
},
{
"epoch": 0.18604229058898267,
"grad_norm": 0.955836256178323,
"learning_rate": 2e-05,
"loss": 0.6910275459289551,
"step": 4100
},
{
"epoch": 0.18649605227334604,
"grad_norm": 2.8875899706209793,
"learning_rate": 2e-05,
"loss": 0.721209716796875,
"step": 4110
},
{
"epoch": 0.1869498139577094,
"grad_norm": 1.3519802172527542,
"learning_rate": 2e-05,
"loss": 0.7201141357421875,
"step": 4120
},
{
"epoch": 0.18740357564207277,
"grad_norm": 1.1465385296144481,
"learning_rate": 2e-05,
"loss": 0.7361567497253418,
"step": 4130
},
{
"epoch": 0.18785733732643617,
"grad_norm": 1.35201264597479,
"learning_rate": 2e-05,
"loss": 0.7152436256408692,
"step": 4140
},
{
"epoch": 0.18831109901079954,
"grad_norm": 2.855369877671521,
"learning_rate": 2e-05,
"loss": 0.732721996307373,
"step": 4150
},
{
"epoch": 0.1887648606951629,
"grad_norm": 1.1504655039133886,
"learning_rate": 2e-05,
"loss": 0.711654806137085,
"step": 4160
},
{
"epoch": 0.18921862237952627,
"grad_norm": 2.435492952972726,
"learning_rate": 2e-05,
"loss": 0.7047530174255371,
"step": 4170
},
{
"epoch": 0.18967238406388964,
"grad_norm": 1.6073142665275684,
"learning_rate": 2e-05,
"loss": 0.7220906734466552,
"step": 4180
},
{
"epoch": 0.190126145748253,
"grad_norm": 5.418513664540345,
"learning_rate": 2e-05,
"loss": 0.7490168571472168,
"step": 4190
},
{
"epoch": 0.19057990743261638,
"grad_norm": 1.5351647225705534,
"learning_rate": 2e-05,
"loss": 0.7134259700775146,
"step": 4200
},
{
"epoch": 0.19103366911697975,
"grad_norm": 2.1996661529885424,
"learning_rate": 2e-05,
"loss": 0.6895709991455078,
"step": 4210
},
{
"epoch": 0.19148743080134314,
"grad_norm": 27.369368463084893,
"learning_rate": 2e-05,
"loss": 0.7323009490966796,
"step": 4220
},
{
"epoch": 0.1919411924857065,
"grad_norm": 1.1069542479218373,
"learning_rate": 2e-05,
"loss": 0.6913928508758544,
"step": 4230
},
{
"epoch": 0.19239495417006988,
"grad_norm": 0.9984480145006956,
"learning_rate": 2e-05,
"loss": 0.7498223781585693,
"step": 4240
},
{
"epoch": 0.19284871585443325,
"grad_norm": 2.7212787049819247,
"learning_rate": 2e-05,
"loss": 0.7058898448944092,
"step": 4250
},
{
"epoch": 0.19330247753879662,
"grad_norm": 1.7354662532878047,
"learning_rate": 2e-05,
"loss": 0.704074764251709,
"step": 4260
},
{
"epoch": 0.19375623922316,
"grad_norm": 5.227313833317236,
"learning_rate": 2e-05,
"loss": 0.6719968318939209,
"step": 4270
},
{
"epoch": 0.19421000090752336,
"grad_norm": 2.253035849395868,
"learning_rate": 2e-05,
"loss": 0.6726584434509277,
"step": 4280
},
{
"epoch": 0.19466376259188675,
"grad_norm": 1.9195847613717556,
"learning_rate": 2e-05,
"loss": 0.7187941551208497,
"step": 4290
},
{
"epoch": 0.19511752427625012,
"grad_norm": 5.06742813791598,
"learning_rate": 2e-05,
"loss": 0.6995963096618653,
"step": 4300
},
{
"epoch": 0.1955712859606135,
"grad_norm": 3.1219867123676783,
"learning_rate": 2e-05,
"loss": 0.7333001613616943,
"step": 4310
},
{
"epoch": 0.19602504764497686,
"grad_norm": 8.171440060022777,
"learning_rate": 2e-05,
"loss": 0.689866018295288,
"step": 4320
},
{
"epoch": 0.19647880932934023,
"grad_norm": 32.66470401435012,
"learning_rate": 2e-05,
"loss": 0.7106449127197265,
"step": 4330
},
{
"epoch": 0.1969325710137036,
"grad_norm": 2.664317300302559,
"learning_rate": 2e-05,
"loss": 0.7009323120117188,
"step": 4340
},
{
"epoch": 0.19738633269806696,
"grad_norm": 3.947837734450377,
"learning_rate": 2e-05,
"loss": 0.7327272415161132,
"step": 4350
},
{
"epoch": 0.19784009438243036,
"grad_norm": 1.4078041437821358,
"learning_rate": 2e-05,
"loss": 0.6748111724853516,
"step": 4360
},
{
"epoch": 0.19829385606679373,
"grad_norm": 1.3395455970025802,
"learning_rate": 2e-05,
"loss": 0.7356967926025391,
"step": 4370
},
{
"epoch": 0.1987476177511571,
"grad_norm": 2.073353287210217,
"learning_rate": 2e-05,
"loss": 0.7055171966552735,
"step": 4380
},
{
"epoch": 0.19920137943552046,
"grad_norm": 1.211049956255314,
"learning_rate": 2e-05,
"loss": 0.6881781578063965,
"step": 4390
},
{
"epoch": 0.19965514111988383,
"grad_norm": 1.1524855810509738,
"learning_rate": 2e-05,
"loss": 0.7047107696533204,
"step": 4400
},
{
"epoch": 0.2001089028042472,
"grad_norm": 1.7427901240259507,
"learning_rate": 2e-05,
"loss": 0.683869457244873,
"step": 4410
},
{
"epoch": 0.20056266448861057,
"grad_norm": 1.377757716899727,
"learning_rate": 2e-05,
"loss": 0.6828576564788819,
"step": 4420
},
{
"epoch": 0.20101642617297397,
"grad_norm": 1.049017045321353,
"learning_rate": 2e-05,
"loss": 0.7124856472015381,
"step": 4430
},
{
"epoch": 0.20147018785733734,
"grad_norm": 1.0254751174088166,
"learning_rate": 2e-05,
"loss": 0.7074912071228028,
"step": 4440
},
{
"epoch": 0.2019239495417007,
"grad_norm": 9.637807908960287,
"learning_rate": 2e-05,
"loss": 0.7565196990966797,
"step": 4450
},
{
"epoch": 0.20237771122606407,
"grad_norm": 1.0043017962801115,
"learning_rate": 2e-05,
"loss": 0.6797826290130615,
"step": 4460
},
{
"epoch": 0.20283147291042744,
"grad_norm": 1.176871506289159,
"learning_rate": 2e-05,
"loss": 0.6929622650146484,
"step": 4470
},
{
"epoch": 0.2032852345947908,
"grad_norm": 5.522800577655823,
"learning_rate": 2e-05,
"loss": 0.6824264049530029,
"step": 4480
},
{
"epoch": 0.20373899627915418,
"grad_norm": 1.3614537335507488,
"learning_rate": 2e-05,
"loss": 0.7002910614013672,
"step": 4490
},
{
"epoch": 0.20419275796351757,
"grad_norm": 2.6673191895352297,
"learning_rate": 2e-05,
"loss": 0.7145938396453857,
"step": 4500
},
{
"epoch": 0.20464651964788094,
"grad_norm": 1.3347810709661898,
"learning_rate": 2e-05,
"loss": 0.6988982200622559,
"step": 4510
},
{
"epoch": 0.2051002813322443,
"grad_norm": 1.2278643663104079,
"learning_rate": 2e-05,
"loss": 0.7082221984863282,
"step": 4520
},
{
"epoch": 0.20555404301660768,
"grad_norm": 1.107773377806117,
"learning_rate": 2e-05,
"loss": 0.7020665168762207,
"step": 4530
},
{
"epoch": 0.20600780470097105,
"grad_norm": 3.8631735434208285,
"learning_rate": 2e-05,
"loss": 0.6998517990112305,
"step": 4540
},
{
"epoch": 0.20646156638533442,
"grad_norm": 1.0646408454737404,
"learning_rate": 2e-05,
"loss": 0.6671296119689941,
"step": 4550
},
{
"epoch": 0.20691532806969778,
"grad_norm": 1.2968846447524345,
"learning_rate": 2e-05,
"loss": 0.6975460052490234,
"step": 4560
},
{
"epoch": 0.20736908975406115,
"grad_norm": 1.2500860468016963,
"learning_rate": 2e-05,
"loss": 0.6920384883880615,
"step": 4570
},
{
"epoch": 0.20782285143842455,
"grad_norm": 2.526585076660618,
"learning_rate": 2e-05,
"loss": 0.7275495052337646,
"step": 4580
},
{
"epoch": 0.20827661312278792,
"grad_norm": 1.826869280378148,
"learning_rate": 2e-05,
"loss": 0.6967020988464355,
"step": 4590
},
{
"epoch": 0.2087303748071513,
"grad_norm": 1.197666025952728,
"learning_rate": 2e-05,
"loss": 0.7026864528656006,
"step": 4600
},
{
"epoch": 0.20918413649151466,
"grad_norm": 2.1162619646227445,
"learning_rate": 2e-05,
"loss": 0.7350880622863769,
"step": 4610
},
{
"epoch": 0.20963789817587802,
"grad_norm": 0.9360778429891393,
"learning_rate": 2e-05,
"loss": 0.7410623550415039,
"step": 4620
},
{
"epoch": 0.2100916598602414,
"grad_norm": 1.184169378243113,
"learning_rate": 2e-05,
"loss": 0.7609868526458741,
"step": 4630
},
{
"epoch": 0.21054542154460476,
"grad_norm": 1.0713710822518603,
"learning_rate": 2e-05,
"loss": 0.7247323513031005,
"step": 4640
},
{
"epoch": 0.21099918322896816,
"grad_norm": 1.3305547090625711,
"learning_rate": 2e-05,
"loss": 0.6840730667114258,
"step": 4650
},
{
"epoch": 0.21145294491333153,
"grad_norm": 1.3917346472940295,
"learning_rate": 2e-05,
"loss": 0.6922812938690186,
"step": 4660
},
{
"epoch": 0.2119067065976949,
"grad_norm": 1.4031498251966887,
"learning_rate": 2e-05,
"loss": 0.6747926712036133,
"step": 4670
},
{
"epoch": 0.21236046828205826,
"grad_norm": 1.2342705730657524,
"learning_rate": 2e-05,
"loss": 0.697972297668457,
"step": 4680
},
{
"epoch": 0.21281422996642163,
"grad_norm": 1.7967651060683705,
"learning_rate": 2e-05,
"loss": 0.7153125762939453,
"step": 4690
},
{
"epoch": 0.213267991650785,
"grad_norm": 1.1416872153149018,
"learning_rate": 2e-05,
"loss": 0.7026151180267334,
"step": 4700
},
{
"epoch": 0.21372175333514837,
"grad_norm": 3.251064960743364,
"learning_rate": 2e-05,
"loss": 0.7278958797454834,
"step": 4710
},
{
"epoch": 0.21417551501951176,
"grad_norm": 2.172677892654505,
"learning_rate": 2e-05,
"loss": 0.688666296005249,
"step": 4720
},
{
"epoch": 0.21462927670387513,
"grad_norm": 1.1460593202443186,
"learning_rate": 2e-05,
"loss": 0.6731379508972168,
"step": 4730
},
{
"epoch": 0.2150830383882385,
"grad_norm": 1.248889845342712,
"learning_rate": 2e-05,
"loss": 0.705271863937378,
"step": 4740
},
{
"epoch": 0.21553680007260187,
"grad_norm": 5.498595536235645,
"learning_rate": 2e-05,
"loss": 0.723629093170166,
"step": 4750
},
{
"epoch": 0.21599056175696524,
"grad_norm": 1.0962041519577148,
"learning_rate": 2e-05,
"loss": 0.7063022136688233,
"step": 4760
},
{
"epoch": 0.2164443234413286,
"grad_norm": 1.2309662992251806,
"learning_rate": 2e-05,
"loss": 0.6802874565124511,
"step": 4770
},
{
"epoch": 0.21689808512569198,
"grad_norm": 1.2726010799073442,
"learning_rate": 2e-05,
"loss": 0.7105306148529053,
"step": 4780
},
{
"epoch": 0.21735184681005537,
"grad_norm": 0.9262030055415884,
"learning_rate": 2e-05,
"loss": 0.6680961608886719,
"step": 4790
},
{
"epoch": 0.21780560849441874,
"grad_norm": 1.2075272466352904,
"learning_rate": 2e-05,
"loss": 0.6859467506408692,
"step": 4800
},
{
"epoch": 0.2182593701787821,
"grad_norm": 1.2386533398195407,
"learning_rate": 2e-05,
"loss": 0.7033780574798584,
"step": 4810
},
{
"epoch": 0.21871313186314548,
"grad_norm": 1.0140778863140478,
"learning_rate": 2e-05,
"loss": 0.6727272033691406,
"step": 4820
},
{
"epoch": 0.21916689354750885,
"grad_norm": 1.4215139270457464,
"learning_rate": 2e-05,
"loss": 0.6806140899658203,
"step": 4830
},
{
"epoch": 0.2196206552318722,
"grad_norm": 1.5176078634404282,
"learning_rate": 2e-05,
"loss": 0.7055531501770019,
"step": 4840
},
{
"epoch": 0.22007441691623558,
"grad_norm": 1.1679352101168905,
"learning_rate": 2e-05,
"loss": 0.6864537715911865,
"step": 4850
},
{
"epoch": 0.22052817860059898,
"grad_norm": 2.8633721580979485,
"learning_rate": 2e-05,
"loss": 0.7271602630615235,
"step": 4860
},
{
"epoch": 0.22098194028496235,
"grad_norm": 3.4849259006369677,
"learning_rate": 2e-05,
"loss": 0.7369964122772217,
"step": 4870
},
{
"epoch": 0.22143570196932572,
"grad_norm": 1.8179425064326888,
"learning_rate": 2e-05,
"loss": 0.7095317840576172,
"step": 4880
},
{
"epoch": 0.22188946365368908,
"grad_norm": 2.9616015396108026,
"learning_rate": 2e-05,
"loss": 0.6696229934692383,
"step": 4890
},
{
"epoch": 0.22234322533805245,
"grad_norm": 1.3778128759705905,
"learning_rate": 2e-05,
"loss": 0.6720298290252685,
"step": 4900
},
{
"epoch": 0.22279698702241582,
"grad_norm": 3.412210539960858,
"learning_rate": 2e-05,
"loss": 0.7047928810119629,
"step": 4910
},
{
"epoch": 0.2232507487067792,
"grad_norm": 1.1599384363613077,
"learning_rate": 2e-05,
"loss": 0.679322338104248,
"step": 4920
},
{
"epoch": 0.22370451039114256,
"grad_norm": 3.1537693015839974,
"learning_rate": 2e-05,
"loss": 0.7178808212280273,
"step": 4930
},
{
"epoch": 0.22415827207550595,
"grad_norm": 1.1360109763405668,
"learning_rate": 2e-05,
"loss": 0.731616735458374,
"step": 4940
},
{
"epoch": 0.22461203375986932,
"grad_norm": 1.2028226670608193,
"learning_rate": 2e-05,
"loss": 0.7333878993988037,
"step": 4950
},
{
"epoch": 0.2250657954442327,
"grad_norm": 4.299549540492807,
"learning_rate": 2e-05,
"loss": 0.6901248931884766,
"step": 4960
},
{
"epoch": 0.22551955712859606,
"grad_norm": 0.9875689724407958,
"learning_rate": 2e-05,
"loss": 0.6959994792938232,
"step": 4970
},
{
"epoch": 0.22597331881295943,
"grad_norm": 1.1120994834812903,
"learning_rate": 2e-05,
"loss": 0.7323190212249756,
"step": 4980
},
{
"epoch": 0.2264270804973228,
"grad_norm": 0.8826964320421593,
"learning_rate": 2e-05,
"loss": 0.660028600692749,
"step": 4990
},
{
"epoch": 0.22688084218168617,
"grad_norm": 1.0440193903081827,
"learning_rate": 2e-05,
"loss": 0.6927733898162842,
"step": 5000
},
{
"epoch": 0.22733460386604956,
"grad_norm": 1.691555422881102,
"learning_rate": 2e-05,
"loss": 0.7029628753662109,
"step": 5010
},
{
"epoch": 0.22778836555041293,
"grad_norm": 1.0344205428775073,
"learning_rate": 2e-05,
"loss": 0.6954968929290771,
"step": 5020
},
{
"epoch": 0.2282421272347763,
"grad_norm": 1.0338848539551273,
"learning_rate": 2e-05,
"loss": 0.6979148864746094,
"step": 5030
},
{
"epoch": 0.22869588891913967,
"grad_norm": 1.279635408493824,
"learning_rate": 2e-05,
"loss": 0.6971482753753662,
"step": 5040
},
{
"epoch": 0.22914965060350304,
"grad_norm": 2.052789784092731,
"learning_rate": 2e-05,
"loss": 0.7221551895141601,
"step": 5050
},
{
"epoch": 0.2296034122878664,
"grad_norm": 5.4430718794689525,
"learning_rate": 2e-05,
"loss": 0.6961151123046875,
"step": 5060
},
{
"epoch": 0.23005717397222977,
"grad_norm": 5.4462313536499,
"learning_rate": 2e-05,
"loss": 0.7519223690032959,
"step": 5070
},
{
"epoch": 0.23051093565659317,
"grad_norm": 2.460852673315715,
"learning_rate": 2e-05,
"loss": 0.6666106700897216,
"step": 5080
},
{
"epoch": 0.23096469734095654,
"grad_norm": 1.9573475864988208,
"learning_rate": 2e-05,
"loss": 0.7317766189575196,
"step": 5090
},
{
"epoch": 0.2314184590253199,
"grad_norm": 1.4614457298710049,
"learning_rate": 2e-05,
"loss": 0.699073600769043,
"step": 5100
},
{
"epoch": 0.23187222070968327,
"grad_norm": 2.179950380633914,
"learning_rate": 2e-05,
"loss": 0.6726521968841552,
"step": 5110
},
{
"epoch": 0.23232598239404664,
"grad_norm": 0.9659552714141566,
"learning_rate": 2e-05,
"loss": 0.6934766292572021,
"step": 5120
},
{
"epoch": 0.23277974407841,
"grad_norm": 1.3476009249911842,
"learning_rate": 2e-05,
"loss": 0.6982085227966308,
"step": 5130
},
{
"epoch": 0.23323350576277338,
"grad_norm": 1.3458662019511773,
"learning_rate": 2e-05,
"loss": 0.7013270378112793,
"step": 5140
},
{
"epoch": 0.23368726744713678,
"grad_norm": 0.9736086189979202,
"learning_rate": 2e-05,
"loss": 0.6897773265838623,
"step": 5150
},
{
"epoch": 0.23414102913150014,
"grad_norm": 0.8622898483806636,
"learning_rate": 2e-05,
"loss": 0.7071221351623536,
"step": 5160
},
{
"epoch": 0.2345947908158635,
"grad_norm": 0.986002446985538,
"learning_rate": 2e-05,
"loss": 0.7530328750610351,
"step": 5170
},
{
"epoch": 0.23504855250022688,
"grad_norm": 23.944927431270106,
"learning_rate": 2e-05,
"loss": 0.7114169120788574,
"step": 5180
},
{
"epoch": 0.23550231418459025,
"grad_norm": 1.3376893951823887,
"learning_rate": 2e-05,
"loss": 0.7076549530029297,
"step": 5190
},
{
"epoch": 0.23595607586895362,
"grad_norm": 2.9129191120482436,
"learning_rate": 2e-05,
"loss": 0.7142559051513672,
"step": 5200
},
{
"epoch": 0.236409837553317,
"grad_norm": 1.0435685509112513,
"learning_rate": 2e-05,
"loss": 0.6851544857025147,
"step": 5210
},
{
"epoch": 0.23686359923768038,
"grad_norm": 1.0472166844507937,
"learning_rate": 2e-05,
"loss": 0.7087789058685303,
"step": 5220
},
{
"epoch": 0.23731736092204375,
"grad_norm": 1.0250703965490964,
"learning_rate": 2e-05,
"loss": 0.7050496578216553,
"step": 5230
},
{
"epoch": 0.23777112260640712,
"grad_norm": 0.9927823437855321,
"learning_rate": 2e-05,
"loss": 0.7455056667327881,
"step": 5240
},
{
"epoch": 0.2382248842907705,
"grad_norm": 1.2528605894234692,
"learning_rate": 2e-05,
"loss": 0.6778655529022217,
"step": 5250
},
{
"epoch": 0.23867864597513386,
"grad_norm": 1.6135191918404508,
"learning_rate": 2e-05,
"loss": 0.7342916488647461,
"step": 5260
},
{
"epoch": 0.23913240765949723,
"grad_norm": 20.780158265081685,
"learning_rate": 2e-05,
"loss": 0.6967748165130615,
"step": 5270
},
{
"epoch": 0.2395861693438606,
"grad_norm": 2.287676275062439,
"learning_rate": 2e-05,
"loss": 0.6897310256958008,
"step": 5280
},
{
"epoch": 0.240039931028224,
"grad_norm": 2.776513468693297,
"learning_rate": 2e-05,
"loss": 0.6728297233581543,
"step": 5290
},
{
"epoch": 0.24049369271258736,
"grad_norm": 1.6269050459577645,
"learning_rate": 2e-05,
"loss": 0.6943089962005615,
"step": 5300
},
{
"epoch": 0.24094745439695073,
"grad_norm": 1.0207451062071664,
"learning_rate": 2e-05,
"loss": 0.7191690444946289,
"step": 5310
},
{
"epoch": 0.2414012160813141,
"grad_norm": 1.824229299150238,
"learning_rate": 2e-05,
"loss": 0.6767844200134278,
"step": 5320
},
{
"epoch": 0.24185497776567746,
"grad_norm": 1.2213809953830987,
"learning_rate": 2e-05,
"loss": 0.6808346748352051,
"step": 5330
},
{
"epoch": 0.24230873945004083,
"grad_norm": 2.2403814470717873,
"learning_rate": 2e-05,
"loss": 0.6660841941833496,
"step": 5340
},
{
"epoch": 0.2427625011344042,
"grad_norm": 1.3580794659047855,
"learning_rate": 1.999194520388485e-05,
"loss": 0.6995076179504395,
"step": 5350
},
{
"epoch": 0.24321626281876757,
"grad_norm": 12.950711471351255,
"learning_rate": 1.9959247303157763e-05,
"loss": 0.6729825973510742,
"step": 5360
},
{
"epoch": 0.24367002450313097,
"grad_norm": 2.0753474395219693,
"learning_rate": 1.990149424459817e-05,
"loss": 0.7218312740325927,
"step": 5370
},
{
"epoch": 0.24412378618749433,
"grad_norm": 0.9844624181474528,
"learning_rate": 1.9818847538648688e-05,
"loss": 0.700545072555542,
"step": 5380
},
{
"epoch": 0.2445775478718577,
"grad_norm": 1.327469277152781,
"learning_rate": 1.971153831256721e-05,
"loss": 0.7063002586364746,
"step": 5390
},
{
"epoch": 0.24503130955622107,
"grad_norm": 5.808958060460233,
"learning_rate": 1.9579866664063435e-05,
"loss": 0.6727919578552246,
"step": 5400
},
{
"epoch": 0.24548507124058444,
"grad_norm": 1.4222807993502444,
"learning_rate": 1.9424200822054782e-05,
"loss": 0.6941525936126709,
"step": 5410
},
{
"epoch": 0.2459388329249478,
"grad_norm": 1.8762814086248882,
"learning_rate": 1.9244976116888626e-05,
"loss": 0.711815595626831,
"step": 5420
},
{
"epoch": 0.24639259460931118,
"grad_norm": 1.354333334832864,
"learning_rate": 1.904269376291071e-05,
"loss": 0.7021115779876709,
"step": 5430
},
{
"epoch": 0.24684635629367457,
"grad_norm": 1.1626523531535664,
"learning_rate": 1.88179194567844e-05,
"loss": 0.679680347442627,
"step": 5440
},
{
"epoch": 0.24730011797803794,
"grad_norm": 6.561971333972244,
"learning_rate": 1.8571281795480632e-05,
"loss": 0.7130425930023193,
"step": 5450
},
{
"epoch": 0.2477538796624013,
"grad_norm": 1.217662089683545,
"learning_rate": 1.8303470518362808e-05,
"loss": 0.6599884033203125,
"step": 5460
},
{
"epoch": 0.24820764134676468,
"grad_norm": 0.9475892650928364,
"learning_rate": 1.801523457828271e-05,
"loss": 0.6993199348449707,
"step": 5470
},
{
"epoch": 0.24866140303112805,
"grad_norm": 0.9601863984631248,
"learning_rate": 1.7707380047081796e-05,
"loss": 0.6852892875671387,
"step": 5480
},
{
"epoch": 0.24911516471549142,
"grad_norm": 1.1235250752571868,
"learning_rate": 1.7380767861355252e-05,
"loss": 0.7044809341430665,
"step": 5490
},
{
"epoch": 0.24956892639985478,
"grad_norm": 1.0077885491038105,
"learning_rate": 1.703631141478299e-05,
"loss": 0.7177154541015625,
"step": 5500
},
{
"epoch": 0.2500226880842182,
"grad_norm": 2.382997200038487,
"learning_rate": 1.6674974003760705e-05,
"loss": 0.7096409797668457,
"step": 5510
},
{
"epoch": 0.2504764497685815,
"grad_norm": 0.9159826545500166,
"learning_rate": 1.6297766133474556e-05,
"loss": 0.671846580505371,
"step": 5520
},
{
"epoch": 0.2509302114529449,
"grad_norm": 3.0409786477401037,
"learning_rate": 1.5905742691953186e-05,
"loss": 0.6969107627868653,
"step": 5530
},
{
"epoch": 0.25138397313730826,
"grad_norm": 1.3842141596912663,
"learning_rate": 1.55e-05,
"loss": 0.680848217010498,
"step": 5540
},
{
"epoch": 0.25183773482167165,
"grad_norm": 10.875469053828501,
"learning_rate": 1.5081672745255862e-05,
"loss": 0.6953877925872802,
"step": 5550
},
{
"epoch": 0.25229149650603505,
"grad_norm": 1.0621859000200418,
"learning_rate": 1.4651930808966206e-05,
"loss": 0.6572223663330078,
"step": 5560
},
{
"epoch": 0.2527452581903984,
"grad_norm": 1.6776124973526256,
"learning_rate": 1.4211975994326848e-05,
"loss": 0.684327220916748,
"step": 5570
},
{
"epoch": 0.2531990198747618,
"grad_norm": 3.6355337932810734,
"learning_rate": 1.376303866555779e-05,
"loss": 0.7440141677856446,
"step": 5580
},
{
"epoch": 0.25365278155912513,
"grad_norm": 0.9600351738335664,
"learning_rate": 1.3306374307104164e-05,
"loss": 0.7563193798065185,
"step": 5590
},
{
"epoch": 0.2541065432434885,
"grad_norm": 1.6375267587917712,
"learning_rate": 1.2843260012586718e-05,
"loss": 0.6717273712158203,
"step": 5600
},
{
"epoch": 0.25456030492785187,
"grad_norm": 1.0745095917709544,
"learning_rate": 1.2374990913320726e-05,
"loss": 0.6807682037353515,
"step": 5610
},
{
"epoch": 0.25501406661221526,
"grad_norm": 1.6201732362530343,
"learning_rate": 1.1902876556391247e-05,
"loss": 0.714454174041748,
"step": 5620
},
{
"epoch": 0.25546782829657866,
"grad_norm": 1.0703093318050196,
"learning_rate": 1.1428237242413685e-05,
"loss": 0.6678834438323975,
"step": 5630
},
{
"epoch": 0.255921589980942,
"grad_norm": 1.2150629537069413,
"learning_rate": 1.095240033322132e-05,
"loss": 0.6891222953796386,
"step": 5640
},
{
"epoch": 0.2563753516653054,
"grad_norm": 11.856539025392681,
"learning_rate": 1.047669653980572e-05,
"loss": 0.6861077785491944,
"step": 5650
},
{
"epoch": 0.25682911334966874,
"grad_norm": 0.8931412033645632,
"learning_rate": 1.0002456200890902e-05,
"loss": 0.662961483001709,
"step": 5660
},
{
"epoch": 0.25728287503403213,
"grad_norm": 2.390051889454472,
"learning_rate": 9.531005562548668e-06,
"loss": 0.6823237419128418,
"step": 5670
},
{
"epoch": 0.2577366367183955,
"grad_norm": 1.0443020757581725,
"learning_rate": 9.06366306925925e-06,
"loss": 0.6859776496887207,
"step": 5680
},
{
"epoch": 0.25819039840275887,
"grad_norm": 1.4203521338174383,
"learning_rate": 8.601735676789686e-06,
"loss": 0.7177210807800293,
"step": 5690
},
{
"epoch": 0.25864416008712227,
"grad_norm": 0.8581315164414361,
"learning_rate": 8.14651519720118e-06,
"loss": 0.6742096900939941,
"step": 5700
},
{
"epoch": 0.2590979217714856,
"grad_norm": 0.9245752076082764,
"learning_rate": 7.699274686206866e-06,
"loss": 0.7056087493896485,
"step": 5710
},
{
"epoch": 0.259551683455849,
"grad_norm": 2.0486524676376594,
"learning_rate": 7.261264882983024e-06,
"loss": 0.663260555267334,
"step": 5720
},
{
"epoch": 0.26000544514021234,
"grad_norm": 4.873855567318575,
"learning_rate": 6.8337107123900386e-06,
"loss": 0.6811215877532959,
"step": 5730
},
{
"epoch": 0.26045920682457574,
"grad_norm": 2.1810371326078837,
"learning_rate": 6.417807859384881e-06,
"loss": 0.6738791465759277,
"step": 5740
},
{
"epoch": 0.2609129685089391,
"grad_norm": 1.3068323593326587,
"learning_rate": 6.014719425205009e-06,
"loss": 0.6841106414794922,
"step": 5750
},
{
"epoch": 0.2613667301933025,
"grad_norm": 1.1251261740134295,
"learning_rate": 5.6255726746749275e-06,
"loss": 0.7019362926483155,
"step": 5760
},
{
"epoch": 0.2618204918776659,
"grad_norm": 0.9964182413989419,
"learning_rate": 5.251455883731728e-06,
"loss": 0.6562312126159668,
"step": 5770
},
{
"epoch": 0.2622742535620292,
"grad_norm": 0.9737757670643322,
"learning_rate": 4.893415295985812e-06,
"loss": 0.6636817932128907,
"step": 5780
},
{
"epoch": 0.2627280152463926,
"grad_norm": 1.2045080283852136,
"learning_rate": 4.552452196827885e-06,
"loss": 0.6575404167175293,
"step": 5790
},
{
"epoch": 0.26318177693075595,
"grad_norm": 8.80624909747458,
"learning_rate": 4.229520113264785e-06,
"loss": 0.6771125793457031,
"step": 5800
},
{
"epoch": 0.26363553861511935,
"grad_norm": 1.0968502820990529,
"learning_rate": 3.925522147314915e-06,
"loss": 0.6726737976074219,
"step": 5810
},
{
"epoch": 0.2640893002994827,
"grad_norm": 1.4519314319569863,
"learning_rate": 3.641308450420675e-06,
"loss": 0.6598453521728516,
"step": 5820
},
{
"epoch": 0.2645430619838461,
"grad_norm": 2.0879351451043062,
"learning_rate": 3.37767384594087e-06,
"loss": 0.6466215133666993,
"step": 5830
},
{
"epoch": 0.2649968236682095,
"grad_norm": 3.10527933306659,
"learning_rate": 3.1353556063719357e-06,
"loss": 0.6572924613952636,
"step": 5840
},
{
"epoch": 0.2654505853525728,
"grad_norm": 1.2323889116224933,
"learning_rate": 2.9150313915141474e-06,
"loss": 0.6518402099609375,
"step": 5850
},
{
"epoch": 0.2659043470369362,
"grad_norm": 0.85550464636182,
"learning_rate": 2.717317353348891e-06,
"loss": 0.6735095024108887,
"step": 5860
},
{
"epoch": 0.26635810872129956,
"grad_norm": 1.0217932865035075,
"learning_rate": 2.5427664129268253e-06,
"loss": 0.6289621353149414,
"step": 5870
},
{
"epoch": 0.26681187040566295,
"grad_norm": 2.1909191238237877,
"learning_rate": 2.391866714085717e-06,
"loss": 0.6425233364105225,
"step": 5880
},
{
"epoch": 0.2672656320900263,
"grad_norm": 1.1470888047889034,
"learning_rate": 2.2650402583222753e-06,
"loss": 0.7085384368896485,
"step": 5890
},
{
"epoch": 0.2677193937743897,
"grad_norm": 1.7414144774236564,
"learning_rate": 2.1626417246356398e-06,
"loss": 0.6326457977294921,
"step": 5900
},
{
"epoch": 0.2681731554587531,
"grad_norm": 0.8499325864885919,
"learning_rate": 2.084957477642894e-06,
"loss": 0.6576657295227051,
"step": 5910
},
{
"epoch": 0.26862691714311643,
"grad_norm": 1.1013515922395056,
"learning_rate": 2.0322047667405147e-06,
"loss": 0.6274625778198242,
"step": 5920
},
{
"epoch": 0.2690806788274798,
"grad_norm": 0.8696399427878706,
"learning_rate": 2.0045311185513342e-06,
"loss": 0.6448920249938965,
"step": 5930
}
],
"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": true
},
"attributes": {}
}
},
"total_flos": 2810342056919040.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}