rqadri's picture
Upload folder using huggingface_hub
ac7bf1c verified
Raw History Blame Contribute Delete
97.3 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.40196782533754666,
"eval_steps": 500,
"global_step": 5597,
"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": 0.8137191520048562,
"learning_rate": 2e-05,
"loss": 0.5525152342660087,
"step": 2550
},
{
"epoch": 0.18385521401896007,
"grad_norm": 0.7485422531957007,
"learning_rate": 2e-05,
"loss": 0.5182879447937012,
"step": 2560
},
{
"epoch": 0.18457339844872164,
"grad_norm": 0.8458986828264131,
"learning_rate": 2e-05,
"loss": 0.5655093193054199,
"step": 2570
},
{
"epoch": 0.18529158287848319,
"grad_norm": 0.8386295808813644,
"learning_rate": 2e-05,
"loss": 0.5635920524597168,
"step": 2580
},
{
"epoch": 0.18600976730824476,
"grad_norm": 1.263649191307002,
"learning_rate": 2e-05,
"loss": 0.5280050277709961,
"step": 2590
},
{
"epoch": 0.18672795173800633,
"grad_norm": 0.6650068518418435,
"learning_rate": 2e-05,
"loss": 0.5367030143737793,
"step": 2600
},
{
"epoch": 0.1874461361677679,
"grad_norm": 0.9820263000565066,
"learning_rate": 2e-05,
"loss": 0.5617701530456543,
"step": 2610
},
{
"epoch": 0.18816432059752944,
"grad_norm": 57.33212038331286,
"learning_rate": 2e-05,
"loss": 0.5331830024719239,
"step": 2620
},
{
"epoch": 0.188882505027291,
"grad_norm": 1.010529867958665,
"learning_rate": 2e-05,
"loss": 0.5588025569915771,
"step": 2630
},
{
"epoch": 0.18960068945705258,
"grad_norm": 2.428568873761783,
"learning_rate": 2e-05,
"loss": 0.5497950553894043,
"step": 2640
},
{
"epoch": 0.19031887388681412,
"grad_norm": 0.9797298582461581,
"learning_rate": 2e-05,
"loss": 0.5596414089202881,
"step": 2650
},
{
"epoch": 0.1910370583165757,
"grad_norm": 1.1976354552542763,
"learning_rate": 2e-05,
"loss": 0.5490626811981201,
"step": 2660
},
{
"epoch": 0.19175524274633726,
"grad_norm": 1.7989529574292977,
"learning_rate": 2e-05,
"loss": 0.545485258102417,
"step": 2670
},
{
"epoch": 0.19247342717609883,
"grad_norm": 1.4264206582276941,
"learning_rate": 2e-05,
"loss": 0.552415418624878,
"step": 2680
},
{
"epoch": 0.19319161160586037,
"grad_norm": 1.6896278687657502,
"learning_rate": 2e-05,
"loss": 0.5572312831878662,
"step": 2690
},
{
"epoch": 0.19390979603562195,
"grad_norm": 1.0240327638926607,
"learning_rate": 2e-05,
"loss": 0.532961368560791,
"step": 2700
},
{
"epoch": 0.19462798046538352,
"grad_norm": 1.6000011559558622,
"learning_rate": 2e-05,
"loss": 0.5521466732025146,
"step": 2710
},
{
"epoch": 0.19534616489514509,
"grad_norm": 2.071118152367986,
"learning_rate": 2e-05,
"loss": 0.5378462791442871,
"step": 2720
},
{
"epoch": 0.19606434932490663,
"grad_norm": 1.2009133478257148,
"learning_rate": 2e-05,
"loss": 0.5343377113342285,
"step": 2730
},
{
"epoch": 0.1967825337546682,
"grad_norm": 2.2291087985430527,
"learning_rate": 2e-05,
"loss": 0.5450372219085693,
"step": 2740
},
{
"epoch": 0.19750071818442977,
"grad_norm": 1.945738285986724,
"learning_rate": 2e-05,
"loss": 0.542167854309082,
"step": 2750
},
{
"epoch": 0.1982189026141913,
"grad_norm": 2.2481238159357004,
"learning_rate": 2e-05,
"loss": 0.5443845272064209,
"step": 2760
},
{
"epoch": 0.19893708704395288,
"grad_norm": 14.705547772660406,
"learning_rate": 2e-05,
"loss": 0.5640571117401123,
"step": 2770
},
{
"epoch": 0.19965527147371445,
"grad_norm": 1.4010674966105274,
"learning_rate": 2e-05,
"loss": 0.5595048904418946,
"step": 2780
},
{
"epoch": 0.20037345590347602,
"grad_norm": 1.6533031936732634,
"learning_rate": 2e-05,
"loss": 0.5488017082214356,
"step": 2790
},
{
"epoch": 0.20109164033323756,
"grad_norm": 0.9101473718897957,
"learning_rate": 2e-05,
"loss": 0.5591398239135742,
"step": 2800
},
{
"epoch": 0.20180982476299913,
"grad_norm": 1.2806154270573438,
"learning_rate": 2e-05,
"loss": 0.5451582908630371,
"step": 2810
},
{
"epoch": 0.2025280091927607,
"grad_norm": 1.096326916624685,
"learning_rate": 2e-05,
"loss": 0.5361979961395263,
"step": 2820
},
{
"epoch": 0.20324619362252228,
"grad_norm": 1.4616941225245348,
"learning_rate": 2e-05,
"loss": 0.523495864868164,
"step": 2830
},
{
"epoch": 0.20396437805228382,
"grad_norm": 4.191564714459604,
"learning_rate": 2e-05,
"loss": 0.5438794612884521,
"step": 2840
},
{
"epoch": 0.2046825624820454,
"grad_norm": 1.7168589035974073,
"learning_rate": 2e-05,
"loss": 0.5668851375579834,
"step": 2850
},
{
"epoch": 0.20540074691180696,
"grad_norm": 0.9229454012230967,
"learning_rate": 2e-05,
"loss": 0.5759838104248047,
"step": 2860
},
{
"epoch": 0.2061189313415685,
"grad_norm": 0.8328522004523823,
"learning_rate": 2e-05,
"loss": 0.5419522762298584,
"step": 2870
},
{
"epoch": 0.20683711577133007,
"grad_norm": 0.9871857471624041,
"learning_rate": 2e-05,
"loss": 0.5363786220550537,
"step": 2880
},
{
"epoch": 0.20755530020109164,
"grad_norm": 3.2487151595898704,
"learning_rate": 2e-05,
"loss": 0.5340115547180175,
"step": 2890
},
{
"epoch": 0.2082734846308532,
"grad_norm": 0.7827799415548224,
"learning_rate": 2e-05,
"loss": 0.5312486171722413,
"step": 2900
},
{
"epoch": 0.20899166906061475,
"grad_norm": 0.9877002420700066,
"learning_rate": 2e-05,
"loss": 0.5470780849456787,
"step": 2910
},
{
"epoch": 0.20970985349037632,
"grad_norm": 1.0995053866218378,
"learning_rate": 2e-05,
"loss": 0.5432174682617188,
"step": 2920
},
{
"epoch": 0.2104280379201379,
"grad_norm": 1.0857728732127427,
"learning_rate": 2e-05,
"loss": 0.5392356872558594,
"step": 2930
},
{
"epoch": 0.21114622234989947,
"grad_norm": 1.9619930189549586,
"learning_rate": 2e-05,
"loss": 0.5705612182617188,
"step": 2940
},
{
"epoch": 0.211864406779661,
"grad_norm": 0.8233017944410961,
"learning_rate": 2e-05,
"loss": 0.5546197414398193,
"step": 2950
},
{
"epoch": 0.21258259120942258,
"grad_norm": 0.9105653047193188,
"learning_rate": 2e-05,
"loss": 0.5508191108703613,
"step": 2960
},
{
"epoch": 0.21330077563918415,
"grad_norm": 0.9269966709109632,
"learning_rate": 2e-05,
"loss": 0.5359363555908203,
"step": 2970
},
{
"epoch": 0.21401896006894572,
"grad_norm": 1.8353878257951093,
"learning_rate": 2e-05,
"loss": 0.5240117073059082,
"step": 2980
},
{
"epoch": 0.21473714449870726,
"grad_norm": 0.9153632235968345,
"learning_rate": 2e-05,
"loss": 0.5855673789978028,
"step": 2990
},
{
"epoch": 0.21545532892846883,
"grad_norm": 2.552566010472789,
"learning_rate": 2e-05,
"loss": 0.5317099571228028,
"step": 3000
},
{
"epoch": 0.2161735133582304,
"grad_norm": 0.7616162209577058,
"learning_rate": 2e-05,
"loss": 0.5220732688903809,
"step": 3010
},
{
"epoch": 0.21689169778799194,
"grad_norm": 0.944591685403448,
"learning_rate": 2e-05,
"loss": 0.5718944549560547,
"step": 3020
},
{
"epoch": 0.21760988221775351,
"grad_norm": 4.068245677494431,
"learning_rate": 2e-05,
"loss": 0.5604113578796387,
"step": 3030
},
{
"epoch": 0.21832806664751508,
"grad_norm": 1.0216540496576307,
"learning_rate": 2e-05,
"loss": 0.5399260520935059,
"step": 3040
},
{
"epoch": 0.21904625107727665,
"grad_norm": 2.1391957462124513,
"learning_rate": 2e-05,
"loss": 0.5619949340820313,
"step": 3050
},
{
"epoch": 0.2197644355070382,
"grad_norm": 7.839074007012571,
"learning_rate": 2e-05,
"loss": 0.5454532146453858,
"step": 3060
},
{
"epoch": 0.22048261993679977,
"grad_norm": 1.0292094224708557,
"learning_rate": 2e-05,
"loss": 0.5300239562988281,
"step": 3070
},
{
"epoch": 0.22120080436656134,
"grad_norm": 1.2148852311315284,
"learning_rate": 2e-05,
"loss": 0.5592824459075928,
"step": 3080
},
{
"epoch": 0.2219189887963229,
"grad_norm": 0.9225477108329542,
"learning_rate": 2e-05,
"loss": 0.5457301139831543,
"step": 3090
},
{
"epoch": 0.22263717322608445,
"grad_norm": 0.8697258595670831,
"learning_rate": 2e-05,
"loss": 0.5454823493957519,
"step": 3100
},
{
"epoch": 0.22335535765584602,
"grad_norm": 5.438955814429222,
"learning_rate": 2e-05,
"loss": 0.5292059898376464,
"step": 3110
},
{
"epoch": 0.2240735420856076,
"grad_norm": 1.0833357383530895,
"learning_rate": 2e-05,
"loss": 0.5511852264404297,
"step": 3120
},
{
"epoch": 0.22479172651536913,
"grad_norm": 0.8984792439415155,
"learning_rate": 2e-05,
"loss": 0.5306966781616211,
"step": 3130
},
{
"epoch": 0.2255099109451307,
"grad_norm": 0.9423509233152327,
"learning_rate": 2e-05,
"loss": 0.5315725803375244,
"step": 3140
},
{
"epoch": 0.22622809537489227,
"grad_norm": 2.4028645931384083,
"learning_rate": 2e-05,
"loss": 0.5435212612152099,
"step": 3150
},
{
"epoch": 0.22694627980465384,
"grad_norm": 1.475845967525974,
"learning_rate": 2e-05,
"loss": 0.5166927814483643,
"step": 3160
},
{
"epoch": 0.2276644642344154,
"grad_norm": 0.6544682011177914,
"learning_rate": 2e-05,
"loss": 0.5181003093719483,
"step": 3170
},
{
"epoch": 0.22838264866417696,
"grad_norm": 1.0973902734239385,
"learning_rate": 2e-05,
"loss": 0.5918588161468505,
"step": 3180
},
{
"epoch": 0.22910083309393853,
"grad_norm": 0.8935409561831856,
"learning_rate": 2e-05,
"loss": 0.5456660270690918,
"step": 3190
},
{
"epoch": 0.2298190175237001,
"grad_norm": 1.1939220525729801,
"learning_rate": 2e-05,
"loss": 0.5290317535400391,
"step": 3200
},
{
"epoch": 0.23053720195346164,
"grad_norm": 1.4729019657064506,
"learning_rate": 2e-05,
"loss": 0.5350526809692383,
"step": 3210
},
{
"epoch": 0.2312553863832232,
"grad_norm": 0.8147013344395122,
"learning_rate": 2e-05,
"loss": 0.5702343940734863,
"step": 3220
},
{
"epoch": 0.23197357081298478,
"grad_norm": 0.9571746233582706,
"learning_rate": 2e-05,
"loss": 0.5327776908874512,
"step": 3230
},
{
"epoch": 0.23269175524274635,
"grad_norm": 0.8061187097411603,
"learning_rate": 2e-05,
"loss": 0.5256548881530761,
"step": 3240
},
{
"epoch": 0.2334099396725079,
"grad_norm": 1.5964531252194205,
"learning_rate": 2e-05,
"loss": 0.5330841064453125,
"step": 3250
},
{
"epoch": 0.23412812410226946,
"grad_norm": 0.8910639877621264,
"learning_rate": 2e-05,
"loss": 0.5282964706420898,
"step": 3260
},
{
"epoch": 0.23484630853203103,
"grad_norm": 0.8365643615103614,
"learning_rate": 2e-05,
"loss": 0.5475823879241943,
"step": 3270
},
{
"epoch": 0.23556449296179258,
"grad_norm": 1.0826561138036208,
"learning_rate": 2e-05,
"loss": 0.5386367797851562,
"step": 3280
},
{
"epoch": 0.23628267739155415,
"grad_norm": 0.8047907032441649,
"learning_rate": 2e-05,
"loss": 0.5200652122497559,
"step": 3290
},
{
"epoch": 0.23700086182131572,
"grad_norm": 1.2596553408647306,
"learning_rate": 2e-05,
"loss": 0.5494902610778809,
"step": 3300
},
{
"epoch": 0.2377190462510773,
"grad_norm": 0.9312337823936057,
"learning_rate": 2e-05,
"loss": 0.5662585258483886,
"step": 3310
},
{
"epoch": 0.23843723068083883,
"grad_norm": 0.7116841214326641,
"learning_rate": 2e-05,
"loss": 0.5465110301971435,
"step": 3320
},
{
"epoch": 0.2391554151106004,
"grad_norm": 1.5807431423736757,
"learning_rate": 2e-05,
"loss": 0.5556153774261474,
"step": 3330
},
{
"epoch": 0.23987359954036197,
"grad_norm": 0.7125361036000691,
"learning_rate": 2e-05,
"loss": 0.5418704986572266,
"step": 3340
},
{
"epoch": 0.24059178397012354,
"grad_norm": 0.949036856277926,
"learning_rate": 2e-05,
"loss": 0.5809379577636719,
"step": 3350
},
{
"epoch": 0.24130996839988508,
"grad_norm": 0.7649138311029305,
"learning_rate": 2e-05,
"loss": 0.5381619930267334,
"step": 3360
},
{
"epoch": 0.24202815282964665,
"grad_norm": 1.0335718019984996,
"learning_rate": 2e-05,
"loss": 0.534903621673584,
"step": 3370
},
{
"epoch": 0.24274633725940822,
"grad_norm": 0.7601362007701609,
"learning_rate": 2e-05,
"loss": 0.541547155380249,
"step": 3380
},
{
"epoch": 0.24346452168916977,
"grad_norm": 1.2909944678819683,
"learning_rate": 2e-05,
"loss": 0.5422227859497071,
"step": 3390
},
{
"epoch": 0.24418270611893134,
"grad_norm": 2.7043297631599246,
"learning_rate": 2e-05,
"loss": 0.5253993988037109,
"step": 3400
},
{
"epoch": 0.2449008905486929,
"grad_norm": 1.6875427780431989,
"learning_rate": 2e-05,
"loss": 0.5760791301727295,
"step": 3410
},
{
"epoch": 0.24561907497845448,
"grad_norm": 3.6344834367174963,
"learning_rate": 2e-05,
"loss": 0.5310137271881104,
"step": 3420
},
{
"epoch": 0.24633725940821602,
"grad_norm": 5.985371030720625,
"learning_rate": 2e-05,
"loss": 0.5414493560791016,
"step": 3430
},
{
"epoch": 0.2470554438379776,
"grad_norm": 1.1163039352095125,
"learning_rate": 2e-05,
"loss": 0.5597766876220703,
"step": 3440
},
{
"epoch": 0.24777362826773916,
"grad_norm": 2.081208345144785,
"learning_rate": 2e-05,
"loss": 0.5317728042602539,
"step": 3450
},
{
"epoch": 0.24849181269750073,
"grad_norm": 2.5443635342349396,
"learning_rate": 2e-05,
"loss": 0.5206192970275879,
"step": 3460
},
{
"epoch": 0.24920999712726227,
"grad_norm": 3.3788478419927532,
"learning_rate": 2e-05,
"loss": 0.5202123641967773,
"step": 3470
},
{
"epoch": 0.24992818155702384,
"grad_norm": 3.0012928738830533,
"learning_rate": 2e-05,
"loss": 0.530025863647461,
"step": 3480
},
{
"epoch": 0.2506463659867854,
"grad_norm": 2.1805517943460417,
"learning_rate": 2e-05,
"loss": 0.5229564666748047,
"step": 3490
},
{
"epoch": 0.251364550416547,
"grad_norm": 1.2494416481651627,
"learning_rate": 2e-05,
"loss": 0.5306306838989258,
"step": 3500
},
{
"epoch": 0.25208273484630855,
"grad_norm": 1.3132397861728369,
"learning_rate": 2e-05,
"loss": 0.5421896934509277,
"step": 3510
},
{
"epoch": 0.25280091927607007,
"grad_norm": 2.0067183252078946,
"learning_rate": 2e-05,
"loss": 0.5233319759368896,
"step": 3520
},
{
"epoch": 0.25351910370583164,
"grad_norm": 3.461715701300714,
"learning_rate": 2e-05,
"loss": 0.5413016319274903,
"step": 3530
},
{
"epoch": 0.2542372881355932,
"grad_norm": 4.944311005364049,
"learning_rate": 2e-05,
"loss": 0.5635467529296875,
"step": 3540
},
{
"epoch": 0.2549554725653548,
"grad_norm": 1.0035627828308078,
"learning_rate": 2e-05,
"loss": 0.5372889995574951,
"step": 3550
},
{
"epoch": 0.25567365699511635,
"grad_norm": 1.5494345748355383,
"learning_rate": 2e-05,
"loss": 0.5421126842498779,
"step": 3560
},
{
"epoch": 0.2563918414248779,
"grad_norm": 1.4255716687651665,
"learning_rate": 2e-05,
"loss": 0.545531177520752,
"step": 3570
},
{
"epoch": 0.2571100258546395,
"grad_norm": 1.0040208201773477,
"learning_rate": 2e-05,
"loss": 0.5218846321105957,
"step": 3580
},
{
"epoch": 0.25782821028440106,
"grad_norm": 2.0175234671803124,
"learning_rate": 2e-05,
"loss": 0.5437562942504883,
"step": 3590
},
{
"epoch": 0.2585463947141626,
"grad_norm": 1.0204550424492405,
"learning_rate": 2e-05,
"loss": 0.5457404136657715,
"step": 3600
},
{
"epoch": 0.25926457914392415,
"grad_norm": 1.157282941042075,
"learning_rate": 2e-05,
"loss": 0.5449418067932129,
"step": 3610
},
{
"epoch": 0.2599827635736857,
"grad_norm": 1.7726889042140002,
"learning_rate": 2e-05,
"loss": 0.5291694641113281,
"step": 3620
},
{
"epoch": 0.2607009480034473,
"grad_norm": 1.2123775554801264,
"learning_rate": 2e-05,
"loss": 0.5180576324462891,
"step": 3630
},
{
"epoch": 0.26141913243320886,
"grad_norm": 3.364579873791338,
"learning_rate": 2e-05,
"loss": 0.5359395980834961,
"step": 3640
},
{
"epoch": 0.2621373168629704,
"grad_norm": 1.302415337588413,
"learning_rate": 2e-05,
"loss": 0.5270327568054199,
"step": 3650
},
{
"epoch": 0.262855501292732,
"grad_norm": 1.9247783430417829,
"learning_rate": 2e-05,
"loss": 0.5509915828704834,
"step": 3660
},
{
"epoch": 0.2635736857224935,
"grad_norm": 2.0236649485917355,
"learning_rate": 2e-05,
"loss": 0.5701016902923584,
"step": 3670
},
{
"epoch": 0.2642918701522551,
"grad_norm": 2.761953593537013,
"learning_rate": 2e-05,
"loss": 0.5551342964172363,
"step": 3680
},
{
"epoch": 0.26501005458201665,
"grad_norm": 1.1643071080332512,
"learning_rate": 2e-05,
"loss": 0.5603652954101562,
"step": 3690
},
{
"epoch": 0.2657282390117782,
"grad_norm": 3.8064230749580283,
"learning_rate": 2e-05,
"loss": 0.555172348022461,
"step": 3700
},
{
"epoch": 0.2664464234415398,
"grad_norm": 2.108256079132036,
"learning_rate": 2e-05,
"loss": 0.5602088451385498,
"step": 3710
},
{
"epoch": 0.26716460787130136,
"grad_norm": 4.8090060875348035,
"learning_rate": 2e-05,
"loss": 0.5268610000610352,
"step": 3720
},
{
"epoch": 0.26788279230106293,
"grad_norm": 11.906837976628907,
"learning_rate": 2e-05,
"loss": 0.5527894496917725,
"step": 3730
},
{
"epoch": 0.2686009767308245,
"grad_norm": 24.394604957331165,
"learning_rate": 2e-05,
"loss": 0.5440473079681396,
"step": 3740
},
{
"epoch": 0.269319161160586,
"grad_norm": 14.268575785205213,
"learning_rate": 2e-05,
"loss": 0.5520382881164551,
"step": 3750
},
{
"epoch": 0.2700373455903476,
"grad_norm": 1.8715318944011055,
"learning_rate": 2e-05,
"loss": 0.5548099040985107,
"step": 3760
},
{
"epoch": 0.27075553002010916,
"grad_norm": 1.3233956016869324,
"learning_rate": 2e-05,
"loss": 0.5696622848510742,
"step": 3770
},
{
"epoch": 0.27147371444987073,
"grad_norm": 2.1153611271898,
"learning_rate": 2e-05,
"loss": 0.5368231773376465,
"step": 3780
},
{
"epoch": 0.2721918988796323,
"grad_norm": 3.4180591319625453,
"learning_rate": 2e-05,
"loss": 0.5424501419067382,
"step": 3790
},
{
"epoch": 0.27291008330939387,
"grad_norm": 1.2561094901094718,
"learning_rate": 2e-05,
"loss": 0.48484387397766116,
"step": 3800
},
{
"epoch": 0.27362826773915544,
"grad_norm": 1.3153456346594474,
"learning_rate": 2e-05,
"loss": 0.5439443111419677,
"step": 3810
},
{
"epoch": 0.27434645216891695,
"grad_norm": 1.1745801668908615,
"learning_rate": 2e-05,
"loss": 0.5262061357498169,
"step": 3820
},
{
"epoch": 0.2750646365986785,
"grad_norm": 0.8624311347459319,
"learning_rate": 2e-05,
"loss": 0.5343884944915771,
"step": 3830
},
{
"epoch": 0.2757828210284401,
"grad_norm": 1.109605104992411,
"learning_rate": 2e-05,
"loss": 0.5506127834320068,
"step": 3840
},
{
"epoch": 0.27650100545820167,
"grad_norm": 1.1537080479350834,
"learning_rate": 2e-05,
"loss": 0.5347657203674316,
"step": 3850
},
{
"epoch": 0.27721918988796324,
"grad_norm": 1.2829687495520894,
"learning_rate": 2e-05,
"loss": 0.5305657386779785,
"step": 3860
},
{
"epoch": 0.2779373743177248,
"grad_norm": 0.9909268980570421,
"learning_rate": 2e-05,
"loss": 0.5470096588134765,
"step": 3870
},
{
"epoch": 0.2786555587474864,
"grad_norm": 3.7169145356207234,
"learning_rate": 2e-05,
"loss": 0.527977180480957,
"step": 3880
},
{
"epoch": 0.2793737431772479,
"grad_norm": 10.90705052951755,
"learning_rate": 2e-05,
"loss": 0.5261877059936524,
"step": 3890
},
{
"epoch": 0.28009192760700946,
"grad_norm": 0.9928164066046667,
"learning_rate": 2e-05,
"loss": 0.5462491989135743,
"step": 3900
},
{
"epoch": 0.28081011203677103,
"grad_norm": 3.3110845903910286,
"learning_rate": 2e-05,
"loss": 0.529015064239502,
"step": 3910
},
{
"epoch": 0.2815282964665326,
"grad_norm": 1.0778076908378762,
"learning_rate": 2e-05,
"loss": 0.5506078243255615,
"step": 3920
},
{
"epoch": 0.28224648089629417,
"grad_norm": 1.1830205823044149,
"learning_rate": 2e-05,
"loss": 0.5306648254394531,
"step": 3930
},
{
"epoch": 0.28296466532605574,
"grad_norm": 3.346211033746781,
"learning_rate": 2e-05,
"loss": 0.521501874923706,
"step": 3940
},
{
"epoch": 0.2836828497558173,
"grad_norm": 1.1684426049709131,
"learning_rate": 2e-05,
"loss": 0.5344061851501465,
"step": 3950
},
{
"epoch": 0.2844010341855789,
"grad_norm": 1.5729556827580824,
"learning_rate": 2e-05,
"loss": 0.5224959373474121,
"step": 3960
},
{
"epoch": 0.2851192186153404,
"grad_norm": 3.346719978096662,
"learning_rate": 2e-05,
"loss": 0.5637972831726075,
"step": 3970
},
{
"epoch": 0.28583740304510197,
"grad_norm": 3.8828663318571026,
"learning_rate": 2e-05,
"loss": 0.5550292015075684,
"step": 3980
},
{
"epoch": 0.28655558747486354,
"grad_norm": 0.8056786165643105,
"learning_rate": 2e-05,
"loss": 0.5798921585083008,
"step": 3990
},
{
"epoch": 0.2872737719046251,
"grad_norm": 4.568679177298758,
"learning_rate": 2e-05,
"loss": 0.5477248668670655,
"step": 4000
},
{
"epoch": 0.2879919563343867,
"grad_norm": 0.7931906689198807,
"learning_rate": 2e-05,
"loss": 0.53890380859375,
"step": 4010
},
{
"epoch": 0.28871014076414825,
"grad_norm": 0.9466268206801044,
"learning_rate": 2e-05,
"loss": 0.5561992645263671,
"step": 4020
},
{
"epoch": 0.2894283251939098,
"grad_norm": 0.9409591371546883,
"learning_rate": 2e-05,
"loss": 0.5654225826263428,
"step": 4030
},
{
"epoch": 0.29014650962367133,
"grad_norm": 0.8526617855655506,
"learning_rate": 2e-05,
"loss": 0.5479049682617188,
"step": 4040
},
{
"epoch": 0.2908646940534329,
"grad_norm": 0.8623694975041977,
"learning_rate": 2e-05,
"loss": 0.5592474937438965,
"step": 4050
},
{
"epoch": 0.2915828784831945,
"grad_norm": 0.9697466994777577,
"learning_rate": 2e-05,
"loss": 0.5225159645080566,
"step": 4060
},
{
"epoch": 0.29230106291295604,
"grad_norm": 1.5124410711764158,
"learning_rate": 2e-05,
"loss": 0.532228660583496,
"step": 4070
},
{
"epoch": 0.2930192473427176,
"grad_norm": 0.876447271347995,
"learning_rate": 2e-05,
"loss": 0.5210915565490722,
"step": 4080
},
{
"epoch": 0.2937374317724792,
"grad_norm": 0.8782796690306948,
"learning_rate": 2e-05,
"loss": 0.5167534351348877,
"step": 4090
},
{
"epoch": 0.29445561620224076,
"grad_norm": 1.0165413549682505,
"learning_rate": 2e-05,
"loss": 0.5146352767944335,
"step": 4100
},
{
"epoch": 0.2951738006320023,
"grad_norm": 1.2412120927299397,
"learning_rate": 2e-05,
"loss": 0.5200153350830078,
"step": 4110
},
{
"epoch": 0.29589198506176384,
"grad_norm": 0.759317713315623,
"learning_rate": 2e-05,
"loss": 0.5243311882019043,
"step": 4120
},
{
"epoch": 0.2966101694915254,
"grad_norm": 0.7688597552315914,
"learning_rate": 2e-05,
"loss": 0.5246932029724121,
"step": 4130
},
{
"epoch": 0.297328353921287,
"grad_norm": 0.8221024219537438,
"learning_rate": 2e-05,
"loss": 0.5325669765472412,
"step": 4140
},
{
"epoch": 0.29804653835104855,
"grad_norm": 4.181610527440594,
"learning_rate": 2e-05,
"loss": 0.5243016242980957,
"step": 4150
},
{
"epoch": 0.2987647227808101,
"grad_norm": 1.14619908037386,
"learning_rate": 2e-05,
"loss": 0.5386734008789062,
"step": 4160
},
{
"epoch": 0.2994829072105717,
"grad_norm": 0.7336771406024591,
"learning_rate": 2e-05,
"loss": 0.5475254058837891,
"step": 4170
},
{
"epoch": 0.30020109164033326,
"grad_norm": 0.8144963221791526,
"learning_rate": 2e-05,
"loss": 0.5383617401123046,
"step": 4180
},
{
"epoch": 0.3009192760700948,
"grad_norm": 0.6807293658810798,
"learning_rate": 2e-05,
"loss": 0.49848508834838867,
"step": 4190
},
{
"epoch": 0.30163746049985635,
"grad_norm": 0.8319765673937244,
"learning_rate": 2e-05,
"loss": 0.5279031753540039,
"step": 4200
},
{
"epoch": 0.3023556449296179,
"grad_norm": 0.7308292287130296,
"learning_rate": 2e-05,
"loss": 0.5480870723724365,
"step": 4210
},
{
"epoch": 0.3030738293593795,
"grad_norm": 0.7245485160597654,
"learning_rate": 2e-05,
"loss": 0.5398597717285156,
"step": 4220
},
{
"epoch": 0.30379201378914106,
"grad_norm": 0.6938652708197344,
"learning_rate": 2e-05,
"loss": 0.535565185546875,
"step": 4230
},
{
"epoch": 0.30451019821890263,
"grad_norm": 1.5661044132536903,
"learning_rate": 2e-05,
"loss": 0.5221335411071777,
"step": 4240
},
{
"epoch": 0.3052283826486642,
"grad_norm": 0.7188791393424939,
"learning_rate": 2e-05,
"loss": 0.5232800483703614,
"step": 4250
},
{
"epoch": 0.3059465670784257,
"grad_norm": 0.9232860721737115,
"learning_rate": 2e-05,
"loss": 0.5546112060546875,
"step": 4260
},
{
"epoch": 0.3066647515081873,
"grad_norm": 0.7980387504616304,
"learning_rate": 2e-05,
"loss": 0.5023956298828125,
"step": 4270
},
{
"epoch": 0.30738293593794885,
"grad_norm": 1.1324128362028514,
"learning_rate": 2e-05,
"loss": 0.5337975978851318,
"step": 4280
},
{
"epoch": 0.3081011203677104,
"grad_norm": 0.8079350872189497,
"learning_rate": 2e-05,
"loss": 0.5285669326782226,
"step": 4290
},
{
"epoch": 0.308819304797472,
"grad_norm": 0.780139719271446,
"learning_rate": 2e-05,
"loss": 0.5422577857971191,
"step": 4300
},
{
"epoch": 0.30953748922723356,
"grad_norm": 0.9078781662253099,
"learning_rate": 2e-05,
"loss": 0.5486246109008789,
"step": 4310
},
{
"epoch": 0.31025567365699513,
"grad_norm": 1.136322453889437,
"learning_rate": 2e-05,
"loss": 0.5188919067382812,
"step": 4320
},
{
"epoch": 0.3109738580867567,
"grad_norm": 0.7250601271434345,
"learning_rate": 2e-05,
"loss": 0.5085612297058105,
"step": 4330
},
{
"epoch": 0.3116920425165182,
"grad_norm": 1.102806118556727,
"learning_rate": 2e-05,
"loss": 0.5358682155609131,
"step": 4340
},
{
"epoch": 0.3124102269462798,
"grad_norm": 0.6439607442931807,
"learning_rate": 2e-05,
"loss": 0.5443156242370606,
"step": 4350
},
{
"epoch": 0.31312841137604136,
"grad_norm": 0.8220346052230498,
"learning_rate": 2e-05,
"loss": 0.5039844036102294,
"step": 4360
},
{
"epoch": 0.31384659580580293,
"grad_norm": 0.6956092309211631,
"learning_rate": 2e-05,
"loss": 0.5142723560333252,
"step": 4370
},
{
"epoch": 0.3145647802355645,
"grad_norm": 0.980336793739194,
"learning_rate": 2e-05,
"loss": 0.5171738147735596,
"step": 4380
},
{
"epoch": 0.31528296466532607,
"grad_norm": 0.677492781126646,
"learning_rate": 2e-05,
"loss": 0.5542737483978272,
"step": 4390
},
{
"epoch": 0.31600114909508764,
"grad_norm": 0.7254282847373281,
"learning_rate": 2e-05,
"loss": 0.5331199645996094,
"step": 4400
},
{
"epoch": 0.31671933352484916,
"grad_norm": 1.2384002962169147,
"learning_rate": 2e-05,
"loss": 0.5397428035736084,
"step": 4410
},
{
"epoch": 0.3174375179546107,
"grad_norm": 0.8616984843210952,
"learning_rate": 2e-05,
"loss": 0.5128665924072265,
"step": 4420
},
{
"epoch": 0.3181557023843723,
"grad_norm": 0.8798407313888418,
"learning_rate": 2e-05,
"loss": 0.5695186614990234,
"step": 4430
},
{
"epoch": 0.31887388681413387,
"grad_norm": 0.752375345775234,
"learning_rate": 2e-05,
"loss": 0.5480000972747803,
"step": 4440
},
{
"epoch": 0.31959207124389544,
"grad_norm": 0.6714903517400259,
"learning_rate": 2e-05,
"loss": 0.5330657958984375,
"step": 4450
},
{
"epoch": 0.320310255673657,
"grad_norm": 0.8014886208045945,
"learning_rate": 2e-05,
"loss": 0.5470023155212402,
"step": 4460
},
{
"epoch": 0.3210284401034186,
"grad_norm": 2.769139896045952,
"learning_rate": 2e-05,
"loss": 0.5314353942871094,
"step": 4470
},
{
"epoch": 0.32174662453318015,
"grad_norm": 0.6920488428651623,
"learning_rate": 2e-05,
"loss": 0.5300350189208984,
"step": 4480
},
{
"epoch": 0.32246480896294166,
"grad_norm": 0.9100167748492725,
"learning_rate": 2e-05,
"loss": 0.5487016677856446,
"step": 4490
},
{
"epoch": 0.32318299339270323,
"grad_norm": 0.895413793344365,
"learning_rate": 2e-05,
"loss": 0.5135931968688965,
"step": 4500
},
{
"epoch": 0.3239011778224648,
"grad_norm": 0.8761216677934326,
"learning_rate": 2e-05,
"loss": 0.5182012557983399,
"step": 4510
},
{
"epoch": 0.3246193622522264,
"grad_norm": 0.6875715837919179,
"learning_rate": 2e-05,
"loss": 0.5375722408294678,
"step": 4520
},
{
"epoch": 0.32533754668198794,
"grad_norm": 0.9467246801297994,
"learning_rate": 2e-05,
"loss": 0.5274314880371094,
"step": 4530
},
{
"epoch": 0.3260557311117495,
"grad_norm": 2.4966362149469288,
"learning_rate": 2e-05,
"loss": 0.5253652572631836,
"step": 4540
},
{
"epoch": 0.3267739155415111,
"grad_norm": 0.9027127429741474,
"learning_rate": 2e-05,
"loss": 0.5107937812805176,
"step": 4550
},
{
"epoch": 0.3274920999712726,
"grad_norm": 1.0302706701334692,
"learning_rate": 2e-05,
"loss": 0.5173007488250733,
"step": 4560
},
{
"epoch": 0.32821028440103417,
"grad_norm": 0.8655105238885601,
"learning_rate": 2e-05,
"loss": 0.5134835243225098,
"step": 4570
},
{
"epoch": 0.32892846883079574,
"grad_norm": 0.8843747358105782,
"learning_rate": 2e-05,
"loss": 0.5232028484344482,
"step": 4580
},
{
"epoch": 0.3296466532605573,
"grad_norm": 0.9324812982026538,
"learning_rate": 2e-05,
"loss": 0.5229804515838623,
"step": 4590
},
{
"epoch": 0.3303648376903189,
"grad_norm": 0.8280836065697736,
"learning_rate": 2e-05,
"loss": 0.5149638652801514,
"step": 4600
},
{
"epoch": 0.33108302212008045,
"grad_norm": 0.7964226439997593,
"learning_rate": 2e-05,
"loss": 0.5428782939910889,
"step": 4610
},
{
"epoch": 0.331801206549842,
"grad_norm": 1.0291422436651445,
"learning_rate": 2e-05,
"loss": 0.5325008392333984,
"step": 4620
},
{
"epoch": 0.33251939097960354,
"grad_norm": 1.2703463158067831,
"learning_rate": 2e-05,
"loss": 0.5268099784851075,
"step": 4630
},
{
"epoch": 0.3332375754093651,
"grad_norm": 0.8085981953561282,
"learning_rate": 2e-05,
"loss": 0.523631477355957,
"step": 4640
},
{
"epoch": 0.3339557598391267,
"grad_norm": 0.7652117796772038,
"learning_rate": 2e-05,
"loss": 0.5244346618652344,
"step": 4650
},
{
"epoch": 0.33467394426888825,
"grad_norm": 0.7630864601816809,
"learning_rate": 2e-05,
"loss": 0.5193476200103759,
"step": 4660
},
{
"epoch": 0.3353921286986498,
"grad_norm": 0.6993476007213558,
"learning_rate": 2e-05,
"loss": 0.5458930969238281,
"step": 4670
},
{
"epoch": 0.3361103131284114,
"grad_norm": 1.1071827966801735,
"learning_rate": 2e-05,
"loss": 0.5582801342010498,
"step": 4680
},
{
"epoch": 0.33682849755817296,
"grad_norm": 1.016677600924547,
"learning_rate": 2e-05,
"loss": 0.5058207988739014,
"step": 4690
},
{
"epoch": 0.3375466819879345,
"grad_norm": 1.2972897369273457,
"learning_rate": 2e-05,
"loss": 0.5481501579284668,
"step": 4700
},
{
"epoch": 0.33826486641769604,
"grad_norm": 0.7564718839488992,
"learning_rate": 2e-05,
"loss": 0.5496013164520264,
"step": 4710
},
{
"epoch": 0.3389830508474576,
"grad_norm": 0.8617166658013237,
"learning_rate": 2e-05,
"loss": 0.5420867919921875,
"step": 4720
},
{
"epoch": 0.3397012352772192,
"grad_norm": 1.7938022936973397,
"learning_rate": 2e-05,
"loss": 0.5397693157196045,
"step": 4730
},
{
"epoch": 0.34041941970698075,
"grad_norm": 0.7552742318995491,
"learning_rate": 2e-05,
"loss": 0.5505653381347656,
"step": 4740
},
{
"epoch": 0.3411376041367423,
"grad_norm": 0.8649411006519387,
"learning_rate": 2e-05,
"loss": 0.5194930076599121,
"step": 4750
},
{
"epoch": 0.3418557885665039,
"grad_norm": 0.7589847446110896,
"learning_rate": 2e-05,
"loss": 0.5183646202087402,
"step": 4760
},
{
"epoch": 0.34257397299626546,
"grad_norm": 0.7993348001820612,
"learning_rate": 2e-05,
"loss": 0.507995319366455,
"step": 4770
},
{
"epoch": 0.343292157426027,
"grad_norm": 0.8597147729918563,
"learning_rate": 2e-05,
"loss": 0.5183065891265869,
"step": 4780
},
{
"epoch": 0.34401034185578855,
"grad_norm": 0.7659204508142745,
"learning_rate": 2e-05,
"loss": 0.5305578231811523,
"step": 4790
},
{
"epoch": 0.3447285262855501,
"grad_norm": 0.9415873647166626,
"learning_rate": 2e-05,
"loss": 0.5384296417236328,
"step": 4800
},
{
"epoch": 0.3454467107153117,
"grad_norm": 0.7565580791373498,
"learning_rate": 2e-05,
"loss": 0.5472804069519043,
"step": 4810
},
{
"epoch": 0.34616489514507326,
"grad_norm": 0.6982311942692283,
"learning_rate": 2e-05,
"loss": 0.5227787971496582,
"step": 4820
},
{
"epoch": 0.34688307957483483,
"grad_norm": 0.6806071049936624,
"learning_rate": 2e-05,
"loss": 0.502852487564087,
"step": 4830
},
{
"epoch": 0.3476012640045964,
"grad_norm": 0.6837855148245949,
"learning_rate": 2e-05,
"loss": 0.5113437175750732,
"step": 4840
},
{
"epoch": 0.34831944843435797,
"grad_norm": 0.7463947130152399,
"learning_rate": 2e-05,
"loss": 0.5150813102722168,
"step": 4850
},
{
"epoch": 0.3490376328641195,
"grad_norm": 1.387103079366797,
"learning_rate": 2e-05,
"loss": 0.5284458637237549,
"step": 4860
},
{
"epoch": 0.34975581729388106,
"grad_norm": 2.105767072263996,
"learning_rate": 2e-05,
"loss": 0.5559632301330566,
"step": 4870
},
{
"epoch": 0.3504740017236426,
"grad_norm": 1.3005516238112327,
"learning_rate": 2e-05,
"loss": 0.5295453548431397,
"step": 4880
},
{
"epoch": 0.3511921861534042,
"grad_norm": 0.8192026159565858,
"learning_rate": 2e-05,
"loss": 0.5568933010101318,
"step": 4890
},
{
"epoch": 0.35191037058316577,
"grad_norm": 1.0967791135442977,
"learning_rate": 2e-05,
"loss": 0.5404356956481934,
"step": 4900
},
{
"epoch": 0.35262855501292734,
"grad_norm": 0.8600343705751858,
"learning_rate": 2e-05,
"loss": 0.5516821861267089,
"step": 4910
},
{
"epoch": 0.3533467394426889,
"grad_norm": 0.8812405283598297,
"learning_rate": 2e-05,
"loss": 0.5511306762695313,
"step": 4920
},
{
"epoch": 0.3540649238724504,
"grad_norm": 0.8898253013696265,
"learning_rate": 2e-05,
"loss": 0.5421256065368653,
"step": 4930
},
{
"epoch": 0.354783108302212,
"grad_norm": 0.7753552574784766,
"learning_rate": 2e-05,
"loss": 0.5207711696624756,
"step": 4940
},
{
"epoch": 0.35550129273197356,
"grad_norm": 0.8196272612662604,
"learning_rate": 2e-05,
"loss": 0.5274653434753418,
"step": 4950
},
{
"epoch": 0.35621947716173513,
"grad_norm": 0.7950338745190594,
"learning_rate": 2e-05,
"loss": 0.5537568092346191,
"step": 4960
},
{
"epoch": 0.3569376615914967,
"grad_norm": 0.9065075356974293,
"learning_rate": 2e-05,
"loss": 0.5296123504638672,
"step": 4970
},
{
"epoch": 0.3576558460212583,
"grad_norm": 1.4889231101858331,
"learning_rate": 2e-05,
"loss": 0.5385382652282715,
"step": 4980
},
{
"epoch": 0.35837403045101984,
"grad_norm": 2.9810240014728016,
"learning_rate": 2e-05,
"loss": 0.5122056484222413,
"step": 4990
},
{
"epoch": 0.35909221488078136,
"grad_norm": 1.0239620796218891,
"learning_rate": 2e-05,
"loss": 0.5193832397460938,
"step": 5000
},
{
"epoch": 0.3598103993105429,
"grad_norm": 1.0813514560466506,
"learning_rate": 2e-05,
"loss": 0.5000585079193115,
"step": 5010
},
{
"epoch": 0.3605285837403045,
"grad_norm": 3.6313651595039795,
"learning_rate": 2e-05,
"loss": 0.552813720703125,
"step": 5020
},
{
"epoch": 0.36124676817006607,
"grad_norm": 3.4464253861451803,
"learning_rate": 2e-05,
"loss": 0.5496349334716797,
"step": 5030
},
{
"epoch": 0.36196495259982764,
"grad_norm": 1.4413205200818051,
"learning_rate": 2e-05,
"loss": 0.5091379165649415,
"step": 5040
},
{
"epoch": 0.3626831370295892,
"grad_norm": 1.2095299773180173,
"learning_rate": 2e-05,
"loss": 0.5228537559509278,
"step": 5050
},
{
"epoch": 0.3634013214593508,
"grad_norm": 1.315218174413597,
"learning_rate": 2e-05,
"loss": 0.5203176021575928,
"step": 5060
},
{
"epoch": 0.36411950588911235,
"grad_norm": 0.8245047599766455,
"learning_rate": 2e-05,
"loss": 0.5189241409301758,
"step": 5070
},
{
"epoch": 0.36483769031887386,
"grad_norm": 0.832552600926099,
"learning_rate": 2e-05,
"loss": 0.529121732711792,
"step": 5080
},
{
"epoch": 0.36555587474863543,
"grad_norm": 0.8671253725370455,
"learning_rate": 1.9999810479633703e-05,
"loss": 0.5332589149475098,
"step": 5090
},
{
"epoch": 0.366274059178397,
"grad_norm": 2.474271903642686,
"learning_rate": 1.997707677020786e-05,
"loss": 0.5437646389007569,
"step": 5100
},
{
"epoch": 0.3669922436081586,
"grad_norm": 1.0572124447219131,
"learning_rate": 1.9916538196124637e-05,
"loss": 0.5545549392700195,
"step": 5110
},
{
"epoch": 0.36771042803792015,
"grad_norm": 0.8268448725243516,
"learning_rate": 1.9818425303863402e-05,
"loss": 0.5480992317199707,
"step": 5120
},
{
"epoch": 0.3684286124676817,
"grad_norm": 6.4552792457201775,
"learning_rate": 1.9683111732582887e-05,
"loss": 0.5488088130950928,
"step": 5130
},
{
"epoch": 0.3691467968974433,
"grad_norm": 0.8832723235171547,
"learning_rate": 1.951111279120706e-05,
"loss": 0.5237607002258301,
"step": 5140
},
{
"epoch": 0.3698649813272048,
"grad_norm": 0.7161551407546988,
"learning_rate": 1.9303083495996028e-05,
"loss": 0.5202799797058105,
"step": 5150
},
{
"epoch": 0.37058316575696637,
"grad_norm": 0.6789244741798717,
"learning_rate": 1.9059816076075482e-05,
"loss": 0.516635513305664,
"step": 5160
},
{
"epoch": 0.37130135018672794,
"grad_norm": 1.0224599652507675,
"learning_rate": 1.8782236956424142e-05,
"loss": 0.5287331581115723,
"step": 5170
},
{
"epoch": 0.3720195346164895,
"grad_norm": 0.6626071055632794,
"learning_rate": 1.8471403229808897e-05,
"loss": 0.5027995586395264,
"step": 5180
},
{
"epoch": 0.3727377190462511,
"grad_norm": 0.7141024429272285,
"learning_rate": 1.8128498631103333e-05,
"loss": 0.5011757850646973,
"step": 5190
},
{
"epoch": 0.37345590347601265,
"grad_norm": 0.9479331921343359,
"learning_rate": 1.7754829029320507e-05,
"loss": 0.5190606117248535,
"step": 5200
},
{
"epoch": 0.3741740879057742,
"grad_norm": 0.8050596491618641,
"learning_rate": 1.7351817454527473e-05,
"loss": 0.5221224784851074,
"step": 5210
},
{
"epoch": 0.3748922723355358,
"grad_norm": 0.8974493994351198,
"learning_rate": 1.6920998678580345e-05,
"loss": 0.5260516166687011,
"step": 5220
},
{
"epoch": 0.3756104567652973,
"grad_norm": 0.7298855828579689,
"learning_rate": 1.6464013370317854e-05,
"loss": 0.5301623821258545,
"step": 5230
},
{
"epoch": 0.3763286411950589,
"grad_norm": 22.805870489924427,
"learning_rate": 1.5982601847471915e-05,
"loss": 0.5027055740356445,
"step": 5240
},
{
"epoch": 0.37704682562482045,
"grad_norm": 0.7174254407063596,
"learning_rate": 1.5478597449089553e-05,
"loss": 0.48400492668151857,
"step": 5250
},
{
"epoch": 0.377765010054582,
"grad_norm": 0.7864566090744084,
"learning_rate": 1.4953919553705725e-05,
"loss": 0.5286859512329102,
"step": 5260
},
{
"epoch": 0.3784831944843436,
"grad_norm": 1.4371153639259329,
"learning_rate": 1.441056626985559e-05,
"loss": 0.5302947521209717,
"step": 5270
},
{
"epoch": 0.37920137891410516,
"grad_norm": 0.7501562644596239,
"learning_rate": 1.3850606826762668e-05,
"loss": 0.5274259567260742,
"step": 5280
},
{
"epoch": 0.37991956334386673,
"grad_norm": 0.7641237043588478,
"learning_rate": 1.3276173694181128e-05,
"loss": 0.5503060817718506,
"step": 5290
},
{
"epoch": 0.38063774777362824,
"grad_norm": 0.6731646505691857,
"learning_rate": 1.2689454461401822e-05,
"loss": 0.5302090167999267,
"step": 5300
},
{
"epoch": 0.3813559322033898,
"grad_norm": 0.8348450893215844,
"learning_rate": 1.2092683506348939e-05,
"loss": 0.4820068359375,
"step": 5310
},
{
"epoch": 0.3820741166331514,
"grad_norm": 0.7767234192966324,
"learning_rate": 1.14881334864935e-05,
"loss": 0.49732160568237305,
"step": 5320
},
{
"epoch": 0.38279230106291295,
"grad_norm": 0.6725539724074863,
"learning_rate": 1.0878106683988481e-05,
"loss": 0.5446413040161133,
"step": 5330
},
{
"epoch": 0.3835104854926745,
"grad_norm": 0.6082174649189866,
"learning_rate": 1.026492623798551e-05,
"loss": 0.4964938163757324,
"step": 5340
},
{
"epoch": 0.3842286699224361,
"grad_norm": 0.888784467154951,
"learning_rate": 9.650927297522743e-06,
"loss": 0.5258268356323242,
"step": 5350
},
{
"epoch": 0.38494685435219766,
"grad_norm": 1.8321946172861412,
"learning_rate": 9.038448128676012e-06,
"loss": 0.5427710533142089,
"step": 5360
},
{
"epoch": 0.3856650387819592,
"grad_norm": 0.9095169733513248,
"learning_rate": 8.42982120983949e-06,
"loss": 0.49727354049682615,
"step": 5370
},
{
"epoch": 0.38638322321172075,
"grad_norm": 0.7410999502181008,
"learning_rate": 7.827364349047303e-06,
"loss": 0.5058761596679687,
"step": 5380
},
{
"epoch": 0.3871014076414823,
"grad_norm": 0.8684719065176256,
"learning_rate": 7.233371857163706e-06,
"loss": 0.5017381191253663,
"step": 5390
},
{
"epoch": 0.3878195920712439,
"grad_norm": 0.8646276820809651,
"learning_rate": 6.650105810556492e-06,
"loss": 0.5159618377685546,
"step": 5400
},
{
"epoch": 0.38853777650100546,
"grad_norm": 0.7107241444937134,
"learning_rate": 6.079787436527681e-06,
"loss": 0.5300826549530029,
"step": 5410
},
{
"epoch": 0.38925596093076703,
"grad_norm": 0.8448464750524096,
"learning_rate": 5.524588654307961e-06,
"loss": 0.5142669200897216,
"step": 5420
},
{
"epoch": 0.3899741453605286,
"grad_norm": 0.6642617294012911,
"learning_rate": 4.986623803828985e-06,
"loss": 0.5289035320281983,
"step": 5430
},
{
"epoch": 0.39069232979029017,
"grad_norm": 1.0224878319471395,
"learning_rate": 4.467941593772423e-06,
"loss": 0.4885417938232422,
"step": 5440
},
{
"epoch": 0.3914105142200517,
"grad_norm": 0.7366024265236012,
"learning_rate": 3.9705172995597606e-06,
"loss": 0.5390846252441406,
"step": 5450
},
{
"epoch": 0.39212869864981326,
"grad_norm": 0.9127526297891573,
"learning_rate": 3.4962452409948346e-06,
"loss": 0.5230414390563964,
"step": 5460
},
{
"epoch": 0.3928468830795748,
"grad_norm": 0.8192724049761325,
"learning_rate": 3.046931568206211e-06,
"loss": 0.5055751800537109,
"step": 5470
},
{
"epoch": 0.3935650675093364,
"grad_norm": 0.6869502617794198,
"learning_rate": 2.624287383362404e-06,
"loss": 0.5020920753479003,
"step": 5480
},
{
"epoch": 0.39428325193909797,
"grad_norm": 0.6558269796596812,
"learning_rate": 2.2299222243541845e-06,
"loss": 0.5411392211914062,
"step": 5490
},
{
"epoch": 0.39500143636885954,
"grad_norm": 0.8058741806747787,
"learning_rate": 1.8653379352598246e-06,
"loss": 0.5336765289306641,
"step": 5500
},
{
"epoch": 0.3957196207986211,
"grad_norm": 0.8868603238138666,
"learning_rate": 1.5319229469361075e-06,
"loss": 0.520848274230957,
"step": 5510
},
{
"epoch": 0.3964378052283826,
"grad_norm": 0.6667796613493242,
"learning_rate": 1.2309469895160917e-06,
"loss": 0.5308838844299316,
"step": 5520
},
{
"epoch": 0.3971559896581442,
"grad_norm": 0.8576035657947603,
"learning_rate": 9.635562569497974e-07,
"loss": 0.5092110633850098,
"step": 5530
},
{
"epoch": 0.39787417408790576,
"grad_norm": 0.6847871783340392,
"learning_rate": 7.307690420024867e-07,
"loss": 0.4752531051635742,
"step": 5540
},
{
"epoch": 0.39859235851766733,
"grad_norm": 0.7194096923536348,
"learning_rate": 5.334718583335942e-07,
"loss": 0.49965767860412597,
"step": 5550
},
{
"epoch": 0.3993105429474289,
"grad_norm": 0.7483011854242718,
"learning_rate": 3.7241606442442763e-07,
"loss": 0.5123620986938476,
"step": 5560
},
{
"epoch": 0.4000287273771905,
"grad_norm": 1.8251975512377843,
"learning_rate": 2.482150022115879e-07,
"loss": 0.5217427730560302,
"step": 5570
},
{
"epoch": 0.40074691180695204,
"grad_norm": 0.7224977292025531,
"learning_rate": 1.6134166132293174e-07,
"loss": 0.5366106986999511,
"step": 5580
},
{
"epoch": 0.4014650962367136,
"grad_norm": 0.6424225207095957,
"learning_rate": 1.1212687781127357e-07,
"loss": 0.5012694358825683,
"step": 5590
},
{
"epoch": 0.40196782533754666,
"step": 5597,
"total_flos": 2650292248117248.0,
"train_loss": 0.047198023267019766,
"train_runtime": 7422.6283,
"train_samples_per_second": 9.049,
"train_steps_per_second": 0.754
}
],
"logging_steps": 10,
"max_steps": 5597,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 636,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 2650292248117248.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}