rqadri's picture
Upload folder using huggingface_hub
9832da8 verified
Raw History Blame Contribute Delete
49.2 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.2008761850043091,
"eval_steps": 500,
"global_step": 2797,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0007181844297615628,
"grad_norm": 175.62312254998636,
"learning_rate": 1.1061797752808989e-06,
"loss": 1.7157390594482422,
"step": 10
},
{
"epoch": 0.0014363688595231256,
"grad_norm": 35.18274256499938,
"learning_rate": 2.2241573033707866e-06,
"loss": 1.1314665794372558,
"step": 20
},
{
"epoch": 0.0021545532892846882,
"grad_norm": 72.02713583586322,
"learning_rate": 3.342134831460675e-06,
"loss": 0.7754706382751465,
"step": 30
},
{
"epoch": 0.002872737719046251,
"grad_norm": 94.14359324942589,
"learning_rate": 4.4601123595505624e-06,
"loss": 0.6444653511047364,
"step": 40
},
{
"epoch": 0.003590922148807814,
"grad_norm": 56.37943749633649,
"learning_rate": 5.5780898876404495e-06,
"loss": 0.5863551616668701,
"step": 50
},
{
"epoch": 0.0043091065785693765,
"grad_norm": 50.395930070118354,
"learning_rate": 6.6960674157303374e-06,
"loss": 0.5876399517059326,
"step": 60
},
{
"epoch": 0.00502729100833094,
"grad_norm": 10.374648240180012,
"learning_rate": 7.814044943820226e-06,
"loss": 0.5711804389953613,
"step": 70
},
{
"epoch": 0.005745475438092502,
"grad_norm": 18.301032576687117,
"learning_rate": 8.932022471910112e-06,
"loss": 0.6154351711273194,
"step": 80
},
{
"epoch": 0.006463659867854065,
"grad_norm": 4.057745910521372,
"learning_rate": 1.005e-05,
"loss": 0.6268198966979981,
"step": 90
},
{
"epoch": 0.007181844297615628,
"grad_norm": 27.638527131076213,
"learning_rate": 1.1167977528089888e-05,
"loss": 0.6075065612792969,
"step": 100
},
{
"epoch": 0.00790002872737719,
"grad_norm": 5.957832731233946,
"learning_rate": 1.2285955056179776e-05,
"loss": 0.5999387264251709,
"step": 110
},
{
"epoch": 0.008618213157138753,
"grad_norm": 8.525671149962339,
"learning_rate": 1.3403932584269664e-05,
"loss": 0.5517457008361817,
"step": 120
},
{
"epoch": 0.009336397586900316,
"grad_norm": 15.665255636675745,
"learning_rate": 1.4521910112359552e-05,
"loss": 0.5644845008850098,
"step": 130
},
{
"epoch": 0.01005458201666188,
"grad_norm": 15.639971766769508,
"learning_rate": 1.563988764044944e-05,
"loss": 0.7366158485412597,
"step": 140
},
{
"epoch": 0.010772766446423441,
"grad_norm": 64.97913496998271,
"learning_rate": 1.6757865168539328e-05,
"loss": 0.6115550518035888,
"step": 150
},
{
"epoch": 0.011490950876185005,
"grad_norm": 59.94539427014833,
"learning_rate": 1.7875842696629214e-05,
"loss": 0.7985194206237793,
"step": 160
},
{
"epoch": 0.012209135305946568,
"grad_norm": 8.629351301874074,
"learning_rate": 1.89938202247191e-05,
"loss": 0.7083045959472656,
"step": 170
},
{
"epoch": 0.01292731973570813,
"grad_norm": 69.06025996795826,
"learning_rate": 2e-05,
"loss": 0.6339869499206543,
"step": 180
},
{
"epoch": 0.013645504165469693,
"grad_norm": 26.037630427264254,
"learning_rate": 2e-05,
"loss": 0.5970722198486328,
"step": 190
},
{
"epoch": 0.014363688595231256,
"grad_norm": 3.7418844806022795,
"learning_rate": 2e-05,
"loss": 0.5959107398986816,
"step": 200
},
{
"epoch": 0.015081873024992818,
"grad_norm": 45.299264079864464,
"learning_rate": 2e-05,
"loss": 0.5934871673583985,
"step": 210
},
{
"epoch": 0.01580005745475438,
"grad_norm": 9.74826111391652,
"learning_rate": 2e-05,
"loss": 0.6289140224456787,
"step": 220
},
{
"epoch": 0.016518241884515943,
"grad_norm": 2.9300194809997895,
"learning_rate": 2e-05,
"loss": 0.5809403896331787,
"step": 230
},
{
"epoch": 0.017236426314277506,
"grad_norm": 54.84408481188558,
"learning_rate": 2e-05,
"loss": 0.5697649955749512,
"step": 240
},
{
"epoch": 0.01795461074403907,
"grad_norm": 259.60423262327515,
"learning_rate": 2e-05,
"loss": 0.6119533061981202,
"step": 250
},
{
"epoch": 0.018672795173800633,
"grad_norm": 9.267214273163608,
"learning_rate": 2e-05,
"loss": 0.651059341430664,
"step": 260
},
{
"epoch": 0.019390979603562196,
"grad_norm": 4.113807630622243,
"learning_rate": 2e-05,
"loss": 0.6351408958435059,
"step": 270
},
{
"epoch": 0.02010916403332376,
"grad_norm": 25.20502044434939,
"learning_rate": 2e-05,
"loss": 0.5994510173797607,
"step": 280
},
{
"epoch": 0.02082734846308532,
"grad_norm": 5.961499873384505,
"learning_rate": 2e-05,
"loss": 0.6210511207580567,
"step": 290
},
{
"epoch": 0.021545532892846882,
"grad_norm": 4.3362120069922945,
"learning_rate": 2e-05,
"loss": 0.6243451595306396,
"step": 300
},
{
"epoch": 0.022263717322608446,
"grad_norm": 4.451056580552574,
"learning_rate": 2e-05,
"loss": 0.6252970695495605,
"step": 310
},
{
"epoch": 0.02298190175237001,
"grad_norm": 18.697593889207074,
"learning_rate": 2e-05,
"loss": 0.6082981109619141,
"step": 320
},
{
"epoch": 0.023700086182131572,
"grad_norm": 4.085643762903478,
"learning_rate": 2e-05,
"loss": 0.6334186553955078,
"step": 330
},
{
"epoch": 0.024418270611893136,
"grad_norm": 2.6889265432904272,
"learning_rate": 2e-05,
"loss": 0.606395673751831,
"step": 340
},
{
"epoch": 0.025136455041654696,
"grad_norm": 4.974541141083005,
"learning_rate": 2e-05,
"loss": 0.5909480571746826,
"step": 350
},
{
"epoch": 0.02585463947141626,
"grad_norm": 43.7940935009098,
"learning_rate": 2e-05,
"loss": 0.6028491020202636,
"step": 360
},
{
"epoch": 0.026572823901177822,
"grad_norm": 12.656136055805545,
"learning_rate": 2e-05,
"loss": 0.5770444869995117,
"step": 370
},
{
"epoch": 0.027291008330939386,
"grad_norm": 2.787735315624888,
"learning_rate": 2e-05,
"loss": 0.601240873336792,
"step": 380
},
{
"epoch": 0.02800919276070095,
"grad_norm": 8.925295369668815,
"learning_rate": 2e-05,
"loss": 0.6033671379089356,
"step": 390
},
{
"epoch": 0.028727377190462512,
"grad_norm": 2.9473827230457927,
"learning_rate": 2e-05,
"loss": 0.5722181320190429,
"step": 400
},
{
"epoch": 0.029445561620224072,
"grad_norm": 3.137535361645301,
"learning_rate": 2e-05,
"loss": 0.5710666656494141,
"step": 410
},
{
"epoch": 0.030163746049985635,
"grad_norm": 2.9468275889813342,
"learning_rate": 2e-05,
"loss": 0.5876998424530029,
"step": 420
},
{
"epoch": 0.0308819304797472,
"grad_norm": 2.9962200372373142,
"learning_rate": 2e-05,
"loss": 0.5958182334899902,
"step": 430
},
{
"epoch": 0.03160011490950876,
"grad_norm": 4.48147209401429,
"learning_rate": 2e-05,
"loss": 0.5968478202819825,
"step": 440
},
{
"epoch": 0.03231829933927032,
"grad_norm": 5.318027567297913,
"learning_rate": 2e-05,
"loss": 0.6289027214050293,
"step": 450
},
{
"epoch": 0.033036483769031885,
"grad_norm": 8.606268913220342,
"learning_rate": 2e-05,
"loss": 0.5964102745056152,
"step": 460
},
{
"epoch": 0.03375466819879345,
"grad_norm": 3.1281299434495837,
"learning_rate": 2e-05,
"loss": 0.6058061122894287,
"step": 470
},
{
"epoch": 0.03447285262855501,
"grad_norm": 4.1439707505248515,
"learning_rate": 2e-05,
"loss": 0.5952574729919433,
"step": 480
},
{
"epoch": 0.035191037058316575,
"grad_norm": 1.2082418063131626,
"learning_rate": 2e-05,
"loss": 0.5965470314025879,
"step": 490
},
{
"epoch": 0.03590922148807814,
"grad_norm": 1.97053507282729,
"learning_rate": 2e-05,
"loss": 0.631954574584961,
"step": 500
},
{
"epoch": 0.0366274059178397,
"grad_norm": 2.8792340786500876,
"learning_rate": 2e-05,
"loss": 0.6188729286193848,
"step": 510
},
{
"epoch": 0.037345590347601265,
"grad_norm": 1.4248895414840272,
"learning_rate": 2e-05,
"loss": 0.5755728721618653,
"step": 520
},
{
"epoch": 0.03806377477736283,
"grad_norm": 10.729751006555638,
"learning_rate": 2e-05,
"loss": 0.5592193603515625,
"step": 530
},
{
"epoch": 0.03878195920712439,
"grad_norm": 6.295780470889749,
"learning_rate": 2e-05,
"loss": 0.5858550071716309,
"step": 540
},
{
"epoch": 0.039500143636885955,
"grad_norm": 5.613715205907156,
"learning_rate": 2e-05,
"loss": 0.6478597164154053,
"step": 550
},
{
"epoch": 0.04021832806664752,
"grad_norm": 2.5638363474088113,
"learning_rate": 2e-05,
"loss": 0.6093330383300781,
"step": 560
},
{
"epoch": 0.040936512496409075,
"grad_norm": 1.608045800845244,
"learning_rate": 2e-05,
"loss": 0.5974137306213378,
"step": 570
},
{
"epoch": 0.04165469692617064,
"grad_norm": 1.3436922071151065,
"learning_rate": 2e-05,
"loss": 0.5949914932250977,
"step": 580
},
{
"epoch": 0.0423728813559322,
"grad_norm": 1.4926769546565009,
"learning_rate": 2e-05,
"loss": 0.5726592540740967,
"step": 590
},
{
"epoch": 0.043091065785693765,
"grad_norm": 2.0317985425865936,
"learning_rate": 2e-05,
"loss": 0.569354248046875,
"step": 600
},
{
"epoch": 0.04380925021545533,
"grad_norm": 1.2071126260231158,
"learning_rate": 2e-05,
"loss": 0.5873561859130859,
"step": 610
},
{
"epoch": 0.04452743464521689,
"grad_norm": 81.77730425521763,
"learning_rate": 2e-05,
"loss": 0.5991431713104248,
"step": 620
},
{
"epoch": 0.045245619074978455,
"grad_norm": 3.1113583484954965,
"learning_rate": 2e-05,
"loss": 0.5880874633789063,
"step": 630
},
{
"epoch": 0.04596380350474002,
"grad_norm": 2.116874681257507,
"learning_rate": 2e-05,
"loss": 0.5938373565673828,
"step": 640
},
{
"epoch": 0.04668198793450158,
"grad_norm": 20.64622489243703,
"learning_rate": 2e-05,
"loss": 0.613460636138916,
"step": 650
},
{
"epoch": 0.047400172364263145,
"grad_norm": 10.809668595672846,
"learning_rate": 2e-05,
"loss": 0.579118299484253,
"step": 660
},
{
"epoch": 0.04811835679402471,
"grad_norm": 8.460108667717218,
"learning_rate": 2e-05,
"loss": 0.571445083618164,
"step": 670
},
{
"epoch": 0.04883654122378627,
"grad_norm": 2.4237687156396506,
"learning_rate": 2e-05,
"loss": 0.5767133235931396,
"step": 680
},
{
"epoch": 0.04955472565354783,
"grad_norm": 1.8961121108699523,
"learning_rate": 2e-05,
"loss": 0.6072338104248047,
"step": 690
},
{
"epoch": 0.05027291008330939,
"grad_norm": 2.0248734200128498,
"learning_rate": 2e-05,
"loss": 0.5928052425384521,
"step": 700
},
{
"epoch": 0.050991094513070954,
"grad_norm": 1.7922158073603092,
"learning_rate": 2e-05,
"loss": 0.5801321983337402,
"step": 710
},
{
"epoch": 0.05170927894283252,
"grad_norm": 1.9541723310096941,
"learning_rate": 2e-05,
"loss": 0.5515483856201172,
"step": 720
},
{
"epoch": 0.05242746337259408,
"grad_norm": 1.1258433200045035,
"learning_rate": 2e-05,
"loss": 0.5742837905883789,
"step": 730
},
{
"epoch": 0.053145647802355644,
"grad_norm": 2.133913640043987,
"learning_rate": 2e-05,
"loss": 0.5683269500732422,
"step": 740
},
{
"epoch": 0.05386383223211721,
"grad_norm": 2.0906475422781963,
"learning_rate": 2e-05,
"loss": 0.5670921325683593,
"step": 750
},
{
"epoch": 0.05458201666187877,
"grad_norm": 4.186492267925719,
"learning_rate": 2e-05,
"loss": 0.590226697921753,
"step": 760
},
{
"epoch": 0.055300201091640334,
"grad_norm": 5.808680786038084,
"learning_rate": 2e-05,
"loss": 0.5766613006591796,
"step": 770
},
{
"epoch": 0.0560183855214019,
"grad_norm": 1.945861940078482,
"learning_rate": 2e-05,
"loss": 0.5599814891815186,
"step": 780
},
{
"epoch": 0.05673656995116346,
"grad_norm": 1.4851684357779684,
"learning_rate": 2e-05,
"loss": 0.5767430305480957,
"step": 790
},
{
"epoch": 0.057454754380925024,
"grad_norm": 1.5127911846708648,
"learning_rate": 2e-05,
"loss": 0.5647124767303466,
"step": 800
},
{
"epoch": 0.05817293881068659,
"grad_norm": 1.909160800633091,
"learning_rate": 2e-05,
"loss": 0.565788745880127,
"step": 810
},
{
"epoch": 0.058891123240448144,
"grad_norm": 1.4125968979208992,
"learning_rate": 2e-05,
"loss": 0.5743461608886719,
"step": 820
},
{
"epoch": 0.05960930767020971,
"grad_norm": 2.387796100293791,
"learning_rate": 2e-05,
"loss": 0.572594165802002,
"step": 830
},
{
"epoch": 0.06032749209997127,
"grad_norm": 1.9203628040448921,
"learning_rate": 2e-05,
"loss": 0.5806607723236084,
"step": 840
},
{
"epoch": 0.061045676529732834,
"grad_norm": 6.326405654102034,
"learning_rate": 2e-05,
"loss": 0.5756526470184327,
"step": 850
},
{
"epoch": 0.0617638609594944,
"grad_norm": 177.16397546897997,
"learning_rate": 2e-05,
"loss": 0.5605918407440186,
"step": 860
},
{
"epoch": 0.06248204538925596,
"grad_norm": 2.8854564417701782,
"learning_rate": 2e-05,
"loss": 0.5636331558227539,
"step": 870
},
{
"epoch": 0.06320022981901752,
"grad_norm": 2.5088047345793925,
"learning_rate": 2e-05,
"loss": 0.5867606163024902,
"step": 880
},
{
"epoch": 0.06391841424877909,
"grad_norm": 1.7473731434882054,
"learning_rate": 2e-05,
"loss": 0.5762314796447754,
"step": 890
},
{
"epoch": 0.06463659867854064,
"grad_norm": 1.9794278765039208,
"learning_rate": 2e-05,
"loss": 0.5760797023773193,
"step": 900
},
{
"epoch": 0.06535478310830221,
"grad_norm": 1.4685265623649186,
"learning_rate": 2e-05,
"loss": 0.5756824493408204,
"step": 910
},
{
"epoch": 0.06607296753806377,
"grad_norm": 3.5706979343374585,
"learning_rate": 2e-05,
"loss": 0.5928406715393066,
"step": 920
},
{
"epoch": 0.06679115196782534,
"grad_norm": 2.410172156848363,
"learning_rate": 2e-05,
"loss": 0.586873197555542,
"step": 930
},
{
"epoch": 0.0675093363975869,
"grad_norm": 1.76346911050291,
"learning_rate": 2e-05,
"loss": 0.5639792919158936,
"step": 940
},
{
"epoch": 0.06822752082734847,
"grad_norm": 1.9137227242149635,
"learning_rate": 2e-05,
"loss": 0.5454784393310547,
"step": 950
},
{
"epoch": 0.06894570525711002,
"grad_norm": 8.304387253857707,
"learning_rate": 2e-05,
"loss": 0.5690406322479248,
"step": 960
},
{
"epoch": 0.0696638896868716,
"grad_norm": 5.601795840800609,
"learning_rate": 2e-05,
"loss": 0.5632596492767334,
"step": 970
},
{
"epoch": 0.07038207411663315,
"grad_norm": 1.2905335866729684,
"learning_rate": 2e-05,
"loss": 0.5677547454833984,
"step": 980
},
{
"epoch": 0.07110025854639472,
"grad_norm": 1.949979768811656,
"learning_rate": 2e-05,
"loss": 0.5854274749755859,
"step": 990
},
{
"epoch": 0.07181844297615628,
"grad_norm": 1.6513016002762253,
"learning_rate": 2e-05,
"loss": 0.5451416015625,
"step": 1000
},
{
"epoch": 0.07253662740591783,
"grad_norm": 1.057713393668006,
"learning_rate": 2e-05,
"loss": 0.5706183433532714,
"step": 1010
},
{
"epoch": 0.0732548118356794,
"grad_norm": 1.19168741304534,
"learning_rate": 2e-05,
"loss": 0.5565749168395996,
"step": 1020
},
{
"epoch": 0.07397299626544096,
"grad_norm": 1.6254916693592154,
"learning_rate": 2e-05,
"loss": 0.5810106277465821,
"step": 1030
},
{
"epoch": 0.07469118069520253,
"grad_norm": 1.5866511613681553,
"learning_rate": 2e-05,
"loss": 0.5619721412658691,
"step": 1040
},
{
"epoch": 0.07540936512496409,
"grad_norm": 1.1498381256400876,
"learning_rate": 2e-05,
"loss": 0.5876872539520264,
"step": 1050
},
{
"epoch": 0.07612754955472566,
"grad_norm": 2.9821157835031227,
"learning_rate": 2e-05,
"loss": 0.5456615447998047,
"step": 1060
},
{
"epoch": 0.07684573398448721,
"grad_norm": 8.097355743313344,
"learning_rate": 2e-05,
"loss": 0.5601712703704834,
"step": 1070
},
{
"epoch": 0.07756391841424878,
"grad_norm": 0.9933021167230655,
"learning_rate": 2e-05,
"loss": 0.5562532424926758,
"step": 1080
},
{
"epoch": 0.07828210284401034,
"grad_norm": 1.8230582482908633,
"learning_rate": 2e-05,
"loss": 0.5808509826660156,
"step": 1090
},
{
"epoch": 0.07900028727377191,
"grad_norm": 4.271370035346493,
"learning_rate": 2e-05,
"loss": 0.5664793968200683,
"step": 1100
},
{
"epoch": 0.07971847170353347,
"grad_norm": 2.5463365356766112,
"learning_rate": 2e-05,
"loss": 0.5739293575286866,
"step": 1110
},
{
"epoch": 0.08043665613329504,
"grad_norm": 2.453175270189324,
"learning_rate": 2e-05,
"loss": 0.5638039588928223,
"step": 1120
},
{
"epoch": 0.0811548405630566,
"grad_norm": 1.5752713585292695,
"learning_rate": 2e-05,
"loss": 0.5842902660369873,
"step": 1130
},
{
"epoch": 0.08187302499281815,
"grad_norm": 1.8751807524553166,
"learning_rate": 2e-05,
"loss": 0.576123046875,
"step": 1140
},
{
"epoch": 0.08259120942257972,
"grad_norm": 1.683663936977947,
"learning_rate": 2e-05,
"loss": 0.569098949432373,
"step": 1150
},
{
"epoch": 0.08330939385234128,
"grad_norm": 0.9580582317849732,
"learning_rate": 2e-05,
"loss": 0.5793169975280762,
"step": 1160
},
{
"epoch": 0.08402757828210285,
"grad_norm": 0.9402295965431912,
"learning_rate": 2e-05,
"loss": 0.5486289024353027,
"step": 1170
},
{
"epoch": 0.0847457627118644,
"grad_norm": 1.1485131505839699,
"learning_rate": 2e-05,
"loss": 0.5426076412200928,
"step": 1180
},
{
"epoch": 0.08546394714162597,
"grad_norm": 1.7549045638499663,
"learning_rate": 2e-05,
"loss": 0.5498751640319824,
"step": 1190
},
{
"epoch": 0.08618213157138753,
"grad_norm": 0.9912225685748786,
"learning_rate": 2e-05,
"loss": 0.5506443023681641,
"step": 1200
},
{
"epoch": 0.0869003160011491,
"grad_norm": 1.3720943673796238,
"learning_rate": 2e-05,
"loss": 0.5797753810882569,
"step": 1210
},
{
"epoch": 0.08761850043091066,
"grad_norm": 1.4192758780321768,
"learning_rate": 2e-05,
"loss": 0.5356600761413575,
"step": 1220
},
{
"epoch": 0.08833668486067223,
"grad_norm": 2.919629967641099,
"learning_rate": 2e-05,
"loss": 0.5717302322387695,
"step": 1230
},
{
"epoch": 0.08905486929043378,
"grad_norm": 1.7267510045028758,
"learning_rate": 2e-05,
"loss": 0.5709489822387696,
"step": 1240
},
{
"epoch": 0.08977305372019534,
"grad_norm": 2.8325078628460765,
"learning_rate": 2e-05,
"loss": 0.5591310501098633,
"step": 1250
},
{
"epoch": 0.09049123814995691,
"grad_norm": 1.986998225545337,
"learning_rate": 2e-05,
"loss": 0.5816420555114746,
"step": 1260
},
{
"epoch": 0.09120942257971847,
"grad_norm": 1.9572581063748238,
"learning_rate": 2e-05,
"loss": 0.5677023887634277,
"step": 1270
},
{
"epoch": 0.09192760700948004,
"grad_norm": 1.6917502403211007,
"learning_rate": 2e-05,
"loss": 0.5681667327880859,
"step": 1280
},
{
"epoch": 0.09264579143924159,
"grad_norm": 11.463447939557085,
"learning_rate": 2e-05,
"loss": 0.5768218040466309,
"step": 1290
},
{
"epoch": 0.09336397586900316,
"grad_norm": 3.8115737877901332,
"learning_rate": 2e-05,
"loss": 0.6065096855163574,
"step": 1300
},
{
"epoch": 0.09408216029876472,
"grad_norm": 2.6413172692463145,
"learning_rate": 2e-05,
"loss": 0.5840601921081543,
"step": 1310
},
{
"epoch": 0.09480034472852629,
"grad_norm": 15.203564161631757,
"learning_rate": 2e-05,
"loss": 0.5864563941955566,
"step": 1320
},
{
"epoch": 0.09551852915828785,
"grad_norm": 2.1523412907056687,
"learning_rate": 2e-05,
"loss": 0.5579684257507325,
"step": 1330
},
{
"epoch": 0.09623671358804942,
"grad_norm": 1.4859209284600976,
"learning_rate": 2e-05,
"loss": 0.550372838973999,
"step": 1340
},
{
"epoch": 0.09695489801781097,
"grad_norm": 1.448109982957294,
"learning_rate": 2e-05,
"loss": 0.5813479900360108,
"step": 1350
},
{
"epoch": 0.09767308244757254,
"grad_norm": 1.3581017917413656,
"learning_rate": 2e-05,
"loss": 0.5561168670654297,
"step": 1360
},
{
"epoch": 0.0983912668773341,
"grad_norm": 6.101417578261775,
"learning_rate": 2e-05,
"loss": 0.5878985404968262,
"step": 1370
},
{
"epoch": 0.09910945130709566,
"grad_norm": 4.598610068508402,
"learning_rate": 2e-05,
"loss": 0.545405101776123,
"step": 1380
},
{
"epoch": 0.09982763573685723,
"grad_norm": 141.15578014813462,
"learning_rate": 2e-05,
"loss": 0.5621461391448974,
"step": 1390
},
{
"epoch": 0.10054582016661878,
"grad_norm": 1.5976483700681585,
"learning_rate": 2e-05,
"loss": 0.565939998626709,
"step": 1400
},
{
"epoch": 0.10126400459638035,
"grad_norm": 1.1844225628657883,
"learning_rate": 2e-05,
"loss": 0.5519757747650147,
"step": 1410
},
{
"epoch": 0.10198218902614191,
"grad_norm": 2.1078647544672737,
"learning_rate": 2e-05,
"loss": 0.5502148628234863,
"step": 1420
},
{
"epoch": 0.10270037345590348,
"grad_norm": 1.1395643545490772,
"learning_rate": 2e-05,
"loss": 0.5794494152069092,
"step": 1430
},
{
"epoch": 0.10341855788566504,
"grad_norm": 1.5152216105283558,
"learning_rate": 2e-05,
"loss": 0.5503892421722412,
"step": 1440
},
{
"epoch": 0.1041367423154266,
"grad_norm": 1.5670725095405607,
"learning_rate": 2e-05,
"loss": 0.5478639125823974,
"step": 1450
},
{
"epoch": 0.10485492674518816,
"grad_norm": 1.4954692560990348,
"learning_rate": 2e-05,
"loss": 0.5685035228729248,
"step": 1460
},
{
"epoch": 0.10557311117494973,
"grad_norm": 1.353128205798565,
"learning_rate": 2e-05,
"loss": 0.5872815132141114,
"step": 1470
},
{
"epoch": 0.10629129560471129,
"grad_norm": 1.7189282039646936,
"learning_rate": 2e-05,
"loss": 0.5590821743011475,
"step": 1480
},
{
"epoch": 0.10700948003447286,
"grad_norm": 1.4249576537485247,
"learning_rate": 2e-05,
"loss": 0.5432151794433594,
"step": 1490
},
{
"epoch": 0.10772766446423442,
"grad_norm": 2.1280600158894867,
"learning_rate": 2e-05,
"loss": 0.5681562423706055,
"step": 1500
},
{
"epoch": 0.10844584889399597,
"grad_norm": 2.4405567792791953,
"learning_rate": 2e-05,
"loss": 0.5866745948791504,
"step": 1510
},
{
"epoch": 0.10916403332375754,
"grad_norm": 3.557452916491406,
"learning_rate": 2e-05,
"loss": 0.5944591522216797,
"step": 1520
},
{
"epoch": 0.1098822177535191,
"grad_norm": 1.5523533701738963,
"learning_rate": 2e-05,
"loss": 0.5890916347503662,
"step": 1530
},
{
"epoch": 0.11060040218328067,
"grad_norm": 2.9322469974839507,
"learning_rate": 2e-05,
"loss": 0.6054721355438233,
"step": 1540
},
{
"epoch": 0.11131858661304223,
"grad_norm": 21.647434412105945,
"learning_rate": 2e-05,
"loss": 0.5661759376525879,
"step": 1550
},
{
"epoch": 0.1120367710428038,
"grad_norm": 1.2627168964462054,
"learning_rate": 2e-05,
"loss": 0.5594186305999755,
"step": 1560
},
{
"epoch": 0.11275495547256535,
"grad_norm": 1.0723943669120632,
"learning_rate": 2e-05,
"loss": 0.5436471939086914,
"step": 1570
},
{
"epoch": 0.11347313990232692,
"grad_norm": 2.4411142355275195,
"learning_rate": 2e-05,
"loss": 0.5494973659515381,
"step": 1580
},
{
"epoch": 0.11419132433208848,
"grad_norm": 1.1969171510706842,
"learning_rate": 2e-05,
"loss": 0.572919225692749,
"step": 1590
},
{
"epoch": 0.11490950876185005,
"grad_norm": 2.483881881946407,
"learning_rate": 2e-05,
"loss": 0.5452249050140381,
"step": 1600
},
{
"epoch": 0.1156276931916116,
"grad_norm": 2.279437521733609,
"learning_rate": 2e-05,
"loss": 0.5241269111633301,
"step": 1610
},
{
"epoch": 0.11634587762137318,
"grad_norm": 3.656360389542672,
"learning_rate": 2e-05,
"loss": 0.5492501735687256,
"step": 1620
},
{
"epoch": 0.11706406205113473,
"grad_norm": 3.4727854166833154,
"learning_rate": 2e-05,
"loss": 0.5788387298583985,
"step": 1630
},
{
"epoch": 0.11778224648089629,
"grad_norm": 1.4914111809005286,
"learning_rate": 2e-05,
"loss": 0.6174117088317871,
"step": 1640
},
{
"epoch": 0.11850043091065786,
"grad_norm": 1.484749826516222,
"learning_rate": 2e-05,
"loss": 0.5739808082580566,
"step": 1650
},
{
"epoch": 0.11921861534041941,
"grad_norm": 2.07472584235867,
"learning_rate": 2e-05,
"loss": 0.607530164718628,
"step": 1660
},
{
"epoch": 0.11993679977018099,
"grad_norm": 1.9705452267412202,
"learning_rate": 2e-05,
"loss": 0.5739453792572021,
"step": 1670
},
{
"epoch": 0.12065498419994254,
"grad_norm": 1.4309591953629426,
"learning_rate": 2e-05,
"loss": 0.5947217464447021,
"step": 1680
},
{
"epoch": 0.12137316862970411,
"grad_norm": 8.183396541359487,
"learning_rate": 2e-05,
"loss": 0.5689792156219482,
"step": 1690
},
{
"epoch": 0.12209135305946567,
"grad_norm": 1.5692583711266774,
"learning_rate": 2e-05,
"loss": 0.5235940456390381,
"step": 1700
},
{
"epoch": 0.12280953748922724,
"grad_norm": 22.072250923049598,
"learning_rate": 2e-05,
"loss": 0.574592399597168,
"step": 1710
},
{
"epoch": 0.1235277219189888,
"grad_norm": 2.1787892001484015,
"learning_rate": 2e-05,
"loss": 0.5768856048583985,
"step": 1720
},
{
"epoch": 0.12424590634875037,
"grad_norm": 1.314265011222165,
"learning_rate": 2e-05,
"loss": 0.5917662620544434,
"step": 1730
},
{
"epoch": 0.12496409077851192,
"grad_norm": 2.561878560439314,
"learning_rate": 2e-05,
"loss": 0.5569777965545655,
"step": 1740
},
{
"epoch": 0.1256822752082735,
"grad_norm": 3.13466478785473,
"learning_rate": 2e-05,
"loss": 0.5714035987854004,
"step": 1750
},
{
"epoch": 0.12640045963803503,
"grad_norm": 1.3850702400017658,
"learning_rate": 2e-05,
"loss": 0.5585092544555664,
"step": 1760
},
{
"epoch": 0.1271186440677966,
"grad_norm": 1.5218161796017777,
"learning_rate": 2e-05,
"loss": 0.557247257232666,
"step": 1770
},
{
"epoch": 0.12783682849755817,
"grad_norm": 2.0961065086106325,
"learning_rate": 2e-05,
"loss": 0.5705356121063232,
"step": 1780
},
{
"epoch": 0.12855501292731975,
"grad_norm": 8.840127197800863,
"learning_rate": 2e-05,
"loss": 0.523622703552246,
"step": 1790
},
{
"epoch": 0.1292731973570813,
"grad_norm": 1.6130732903200282,
"learning_rate": 2e-05,
"loss": 0.5404885768890381,
"step": 1800
},
{
"epoch": 0.12999138178684286,
"grad_norm": 3.3209600372002086,
"learning_rate": 2e-05,
"loss": 0.5290946960449219,
"step": 1810
},
{
"epoch": 0.13070956621660443,
"grad_norm": 1.7926966358429328,
"learning_rate": 2e-05,
"loss": 0.5992692470550537,
"step": 1820
},
{
"epoch": 0.131427750646366,
"grad_norm": 2.3058228432645897,
"learning_rate": 2e-05,
"loss": 0.5432312965393067,
"step": 1830
},
{
"epoch": 0.13214593507612754,
"grad_norm": 2.001620000676492,
"learning_rate": 2e-05,
"loss": 0.5786615371704101,
"step": 1840
},
{
"epoch": 0.1328641195058891,
"grad_norm": 1.4184231189324574,
"learning_rate": 2e-05,
"loss": 0.5465398788452148,
"step": 1850
},
{
"epoch": 0.13358230393565068,
"grad_norm": 1.7730327835422113,
"learning_rate": 2e-05,
"loss": 0.5668454170227051,
"step": 1860
},
{
"epoch": 0.13430048836541225,
"grad_norm": 1.5629284623062545,
"learning_rate": 2e-05,
"loss": 0.6058121681213379,
"step": 1870
},
{
"epoch": 0.1350186727951738,
"grad_norm": 4.630139619066831,
"learning_rate": 2e-05,
"loss": 0.5537769317626953,
"step": 1880
},
{
"epoch": 0.13573685722493536,
"grad_norm": 3.5059792014239473,
"learning_rate": 2e-05,
"loss": 0.5759965419769287,
"step": 1890
},
{
"epoch": 0.13645504165469693,
"grad_norm": 1.329974685954622,
"learning_rate": 2e-05,
"loss": 0.5549535274505615,
"step": 1900
},
{
"epoch": 0.13717322608445848,
"grad_norm": 6.710102458220659,
"learning_rate": 2e-05,
"loss": 0.5439998149871826,
"step": 1910
},
{
"epoch": 0.13789141051422005,
"grad_norm": 4.374063390421496,
"learning_rate": 2e-05,
"loss": 0.5546764373779297,
"step": 1920
},
{
"epoch": 0.13860959494398162,
"grad_norm": 2.6173407845738166,
"learning_rate": 2e-05,
"loss": 0.5726741790771485,
"step": 1930
},
{
"epoch": 0.1393277793737432,
"grad_norm": 4.304682258340256,
"learning_rate": 2e-05,
"loss": 0.5658658504486084,
"step": 1940
},
{
"epoch": 0.14004596380350473,
"grad_norm": 3.3504816392214782,
"learning_rate": 2e-05,
"loss": 0.5655290126800537,
"step": 1950
},
{
"epoch": 0.1407641482332663,
"grad_norm": 1.838294224798907,
"learning_rate": 2e-05,
"loss": 0.5578551769256592,
"step": 1960
},
{
"epoch": 0.14148233266302787,
"grad_norm": 3.008317667214241,
"learning_rate": 2e-05,
"loss": 0.5349006652832031,
"step": 1970
},
{
"epoch": 0.14220051709278944,
"grad_norm": 3.013445284817389,
"learning_rate": 2e-05,
"loss": 0.5707347393035889,
"step": 1980
},
{
"epoch": 0.14291870152255098,
"grad_norm": 1.7167470080649117,
"learning_rate": 2e-05,
"loss": 0.5561907768249512,
"step": 1990
},
{
"epoch": 0.14363688595231255,
"grad_norm": 2.2653225742012615,
"learning_rate": 2e-05,
"loss": 0.5378569602966309,
"step": 2000
},
{
"epoch": 0.14435507038207412,
"grad_norm": 3.3495014969229766,
"learning_rate": 2e-05,
"loss": 0.5674232482910156,
"step": 2010
},
{
"epoch": 0.14507325481183567,
"grad_norm": 1.7093094301537985,
"learning_rate": 2e-05,
"loss": 0.5741843223571778,
"step": 2020
},
{
"epoch": 0.14579143924159724,
"grad_norm": 1.5010655332169438,
"learning_rate": 2e-05,
"loss": 0.5695930004119873,
"step": 2030
},
{
"epoch": 0.1465096236713588,
"grad_norm": 1.5521694994775619,
"learning_rate": 2e-05,
"loss": 0.5523951530456543,
"step": 2040
},
{
"epoch": 0.14722780810112038,
"grad_norm": 3.0404224380491347,
"learning_rate": 2e-05,
"loss": 0.5805356502532959,
"step": 2050
},
{
"epoch": 0.14794599253088192,
"grad_norm": 1.349791102810318,
"learning_rate": 2e-05,
"loss": 0.5671909332275391,
"step": 2060
},
{
"epoch": 0.1486641769606435,
"grad_norm": 3.6695671293215697,
"learning_rate": 2e-05,
"loss": 0.5312755584716797,
"step": 2070
},
{
"epoch": 0.14938236139040506,
"grad_norm": 4.911695581931882,
"learning_rate": 2e-05,
"loss": 0.5726969718933106,
"step": 2080
},
{
"epoch": 0.15010054582016663,
"grad_norm": 8.09684298774792,
"learning_rate": 2e-05,
"loss": 0.5571673393249512,
"step": 2090
},
{
"epoch": 0.15081873024992817,
"grad_norm": 2.2745544981394143,
"learning_rate": 2e-05,
"loss": 0.5517287254333496,
"step": 2100
},
{
"epoch": 0.15153691467968974,
"grad_norm": 1.419765240616946,
"learning_rate": 2e-05,
"loss": 0.5615994930267334,
"step": 2110
},
{
"epoch": 0.15225509910945131,
"grad_norm": 3.7635018424594864,
"learning_rate": 2e-05,
"loss": 0.566252613067627,
"step": 2120
},
{
"epoch": 0.15297328353921286,
"grad_norm": 2.765211749203788,
"learning_rate": 2e-05,
"loss": 0.5662132263183594,
"step": 2130
},
{
"epoch": 0.15369146796897443,
"grad_norm": 19.961112859254513,
"learning_rate": 2e-05,
"loss": 0.5998068809509277,
"step": 2140
},
{
"epoch": 0.154409652398736,
"grad_norm": 1.7784559217289087,
"learning_rate": 2e-05,
"loss": 0.564358139038086,
"step": 2150
},
{
"epoch": 0.15512783682849757,
"grad_norm": 1.3875792261494675,
"learning_rate": 2e-05,
"loss": 0.5419380187988281,
"step": 2160
},
{
"epoch": 0.1558460212582591,
"grad_norm": 2.022411553935775,
"learning_rate": 2e-05,
"loss": 0.5899745941162109,
"step": 2170
},
{
"epoch": 0.15656420568802068,
"grad_norm": 2.272563024973418,
"learning_rate": 2e-05,
"loss": 0.55307035446167,
"step": 2180
},
{
"epoch": 0.15728239011778225,
"grad_norm": 1.9234961093353966,
"learning_rate": 2e-05,
"loss": 0.6007951259613037,
"step": 2190
},
{
"epoch": 0.15800057454754382,
"grad_norm": 1.256053229715644,
"learning_rate": 2e-05,
"loss": 0.5546849250793457,
"step": 2200
},
{
"epoch": 0.15871875897730536,
"grad_norm": 1.2335218202043055,
"learning_rate": 2e-05,
"loss": 0.5476455688476562,
"step": 2210
},
{
"epoch": 0.15943694340706693,
"grad_norm": 2.8401539385475116,
"learning_rate": 2e-05,
"loss": 0.5543958187103272,
"step": 2220
},
{
"epoch": 0.1601551278368285,
"grad_norm": 2.197748282778052,
"learning_rate": 2e-05,
"loss": 0.5589810848236084,
"step": 2230
},
{
"epoch": 0.16087331226659007,
"grad_norm": 3.3213977186166805,
"learning_rate": 2e-05,
"loss": 0.5354520320892334,
"step": 2240
},
{
"epoch": 0.16159149669635162,
"grad_norm": 1.445633989615147,
"learning_rate": 2e-05,
"loss": 0.5722284317016602,
"step": 2250
},
{
"epoch": 0.1623096811261132,
"grad_norm": 7.116743865111077,
"learning_rate": 2e-05,
"loss": 0.5650218963623047,
"step": 2260
},
{
"epoch": 0.16302786555587476,
"grad_norm": 1.716127091438019,
"learning_rate": 2e-05,
"loss": 0.5529635429382325,
"step": 2270
},
{
"epoch": 0.1637460499856363,
"grad_norm": 1.055030692862526,
"learning_rate": 2e-05,
"loss": 0.5603778839111329,
"step": 2280
},
{
"epoch": 0.16446423441539787,
"grad_norm": 1.342803947507129,
"learning_rate": 2e-05,
"loss": 0.5632447242736817,
"step": 2290
},
{
"epoch": 0.16518241884515944,
"grad_norm": 1.054694302356739,
"learning_rate": 2e-05,
"loss": 0.5708892822265625,
"step": 2300
},
{
"epoch": 0.165900603274921,
"grad_norm": 1.1307530423263086,
"learning_rate": 2e-05,
"loss": 0.5213704109191895,
"step": 2310
},
{
"epoch": 0.16661878770468255,
"grad_norm": 1.0820602105055892,
"learning_rate": 2e-05,
"loss": 0.5062259197235107,
"step": 2320
},
{
"epoch": 0.16733697213444412,
"grad_norm": 1.1270089576544027,
"learning_rate": 2e-05,
"loss": 0.5432156562805176,
"step": 2330
},
{
"epoch": 0.1680551565642057,
"grad_norm": 0.9920832286718221,
"learning_rate": 2e-05,
"loss": 0.5382828712463379,
"step": 2340
},
{
"epoch": 0.16877334099396726,
"grad_norm": 1.1726405888684055,
"learning_rate": 2e-05,
"loss": 0.5809371948242188,
"step": 2350
},
{
"epoch": 0.1694915254237288,
"grad_norm": 0.9340362083469751,
"learning_rate": 2e-05,
"loss": 0.549803352355957,
"step": 2360
},
{
"epoch": 0.17020970985349038,
"grad_norm": 0.938566558796553,
"learning_rate": 2e-05,
"loss": 0.5452708721160888,
"step": 2370
},
{
"epoch": 0.17092789428325195,
"grad_norm": 0.7822121686397506,
"learning_rate": 2e-05,
"loss": 0.573302173614502,
"step": 2380
},
{
"epoch": 0.1716460787130135,
"grad_norm": 0.8584393263957636,
"learning_rate": 2e-05,
"loss": 0.5399321556091309,
"step": 2390
},
{
"epoch": 0.17236426314277506,
"grad_norm": 1.0102345883460102,
"learning_rate": 2e-05,
"loss": 0.5162699222564697,
"step": 2400
},
{
"epoch": 0.17308244757253663,
"grad_norm": 0.9712142063036817,
"learning_rate": 2e-05,
"loss": 0.546810531616211,
"step": 2410
},
{
"epoch": 0.1738006320022982,
"grad_norm": 1.092566228410939,
"learning_rate": 2e-05,
"loss": 0.5729592800140381,
"step": 2420
},
{
"epoch": 0.17451881643205974,
"grad_norm": 0.7737223185881741,
"learning_rate": 2e-05,
"loss": 0.5603939056396484,
"step": 2430
},
{
"epoch": 0.1752370008618213,
"grad_norm": 1.1017251656675449,
"learning_rate": 2e-05,
"loss": 0.5336654663085938,
"step": 2440
},
{
"epoch": 0.17595518529158288,
"grad_norm": 1.881306561126187,
"learning_rate": 2e-05,
"loss": 0.5107797622680664,
"step": 2450
},
{
"epoch": 0.17667336972134445,
"grad_norm": 0.8552534631118969,
"learning_rate": 2e-05,
"loss": 0.5397173881530761,
"step": 2460
},
{
"epoch": 0.177391554151106,
"grad_norm": 1.1626898542148933,
"learning_rate": 2e-05,
"loss": 0.5552604675292969,
"step": 2470
},
{
"epoch": 0.17810973858086757,
"grad_norm": 0.7392869255775882,
"learning_rate": 2e-05,
"loss": 0.5463393211364747,
"step": 2480
},
{
"epoch": 0.17882792301062914,
"grad_norm": 0.8337215511497923,
"learning_rate": 2e-05,
"loss": 0.5500159740447998,
"step": 2490
},
{
"epoch": 0.17954610744039068,
"grad_norm": 1.2250992404021477,
"learning_rate": 2e-05,
"loss": 0.5643744468688965,
"step": 2500
},
{
"epoch": 0.18026429187015225,
"grad_norm": 1.9461609461436875,
"learning_rate": 2e-05,
"loss": 0.565704345703125,
"step": 2510
},
{
"epoch": 0.18098247629991382,
"grad_norm": 2.6973002486791806,
"learning_rate": 2e-05,
"loss": 0.5357652187347413,
"step": 2520
},
{
"epoch": 0.1817006607296754,
"grad_norm": 1.219804225085043,
"learning_rate": 2e-05,
"loss": 0.5559783458709717,
"step": 2530
},
{
"epoch": 0.18241884515943693,
"grad_norm": 1.2921261695545436,
"learning_rate": 2e-05,
"loss": 0.5474600791931152,
"step": 2540
},
{
"epoch": 0.1831370295891985,
"grad_norm": 0.9360248915766971,
"learning_rate": 1.997261400182901e-05,
"loss": 0.5525213990892682,
"step": 2550
},
{
"epoch": 0.18385521401896007,
"grad_norm": 0.7132455157008145,
"learning_rate": 1.9805800746810423e-05,
"loss": 0.5178929328918457,
"step": 2560
},
{
"epoch": 0.18457339844872164,
"grad_norm": 1.3860288183277505,
"learning_rate": 1.9489934220110274e-05,
"loss": 0.5651967525482178,
"step": 2570
},
{
"epoch": 0.18529158287848319,
"grad_norm": 1.0773710559684646,
"learning_rate": 1.9029840364820392e-05,
"loss": 0.5613128185272217,
"step": 2580
},
{
"epoch": 0.18600976730824476,
"grad_norm": 1.0893172560174817,
"learning_rate": 1.8432548690479908e-05,
"loss": 0.5255576610565186,
"step": 2590
},
{
"epoch": 0.18672795173800633,
"grad_norm": 0.8548975503190002,
"learning_rate": 1.7707184873239772e-05,
"loss": 0.5357662200927734,
"step": 2600
},
{
"epoch": 0.1874461361677679,
"grad_norm": 1.8089525554439896,
"learning_rate": 1.6864831329902565e-05,
"loss": 0.5596217155456543,
"step": 2610
},
{
"epoch": 0.18816432059752944,
"grad_norm": 1.8086804118521456,
"learning_rate": 1.591835789604667e-05,
"loss": 0.5309314250946044,
"step": 2620
},
{
"epoch": 0.188882505027291,
"grad_norm": 3.3177584130017492,
"learning_rate": 1.4882225195206459e-05,
"loss": 0.5537728786468505,
"step": 2630
},
{
"epoch": 0.18960068945705258,
"grad_norm": 1.5730670133706086,
"learning_rate": 1.3772263703317695e-05,
"loss": 0.5419403076171875,
"step": 2640
},
{
"epoch": 0.19031887388681412,
"grad_norm": 1.4481745706842108,
"learning_rate": 1.2605431883975547e-05,
"loss": 0.5550278663635254,
"step": 2650
},
{
"epoch": 0.1910370583165757,
"grad_norm": 0.9496183443198778,
"learning_rate": 1.1399557089817574e-05,
"loss": 0.5444825649261474,
"step": 2660
},
{
"epoch": 0.19175524274633726,
"grad_norm": 4.222941220284988,
"learning_rate": 1.0173063188650463e-05,
"loss": 0.5412864685058594,
"step": 2670
},
{
"epoch": 0.19247342717609883,
"grad_norm": 5.253318465543954,
"learning_rate": 8.944689075764396e-06,
"loss": 0.5456050872802735,
"step": 2680
},
{
"epoch": 0.19319161160586037,
"grad_norm": 0.9111342540296383,
"learning_rate": 7.73320237312362e-06,
"loss": 0.5468400955200196,
"step": 2690
},
{
"epoch": 0.19390979603562195,
"grad_norm": 0.8304532982880024,
"learning_rate": 6.557112689658629e-06,
"loss": 0.5211694240570068,
"step": 2700
},
{
"epoch": 0.19462798046538352,
"grad_norm": 0.9417548584277616,
"learning_rate": 5.434388823591213e-06,
"loss": 0.539794635772705,
"step": 2710
},
{
"epoch": 0.19534616489514509,
"grad_norm": 0.9364997135575152,
"learning_rate": 4.382184227495424e-06,
"loss": 0.5243682861328125,
"step": 2720
},
{
"epoch": 0.19606434932490663,
"grad_norm": 0.9086225580589578,
"learning_rate": 3.4165749305560203e-06,
"loss": 0.5202994346618652,
"step": 2730
},
{
"epoch": 0.1967825337546682,
"grad_norm": 1.8965570428273997,
"learning_rate": 2.5523139221595556e-06,
"loss": 0.5278103351593018,
"step": 2740
},
{
"epoch": 0.19750071818442977,
"grad_norm": 0.8511964964429284,
"learning_rate": 1.8026057494499821e-06,
"loss": 0.5267528533935547,
"step": 2750
},
{
"epoch": 0.1982189026141913,
"grad_norm": 0.7914798757409518,
"learning_rate": 1.178904772643223e-06,
"loss": 0.5294023990631104,
"step": 2760
},
{
"epoch": 0.19893708704395288,
"grad_norm": 0.7573600325272046,
"learning_rate": 6.90740160441754e-07,
"loss": 0.542425537109375,
"step": 2770
},
{
"epoch": 0.19965527147371445,
"grad_norm": 1.549889817778262,
"learning_rate": 3.45570299343632e-07,
"loss": 0.5377500057220459,
"step": 2780
},
{
"epoch": 0.20037345590347602,
"grad_norm": 1.0279344434076476,
"learning_rate": 1.486688412425152e-07,
"loss": 0.5264135360717773,
"step": 2790
},
{
"epoch": 0.2008761850043091,
"step": 2797,
"total_flos": 1324435843842048.0,
"train_loss": 0.04869218967111783,
"train_runtime": 4220.8156,
"train_samples_per_second": 7.952,
"train_steps_per_second": 0.663
}
],
"logging_steps": 10,
"max_steps": 2797,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 2543,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1324435843842048.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}