rqadri's picture
Upload folder using huggingface_hub
c1ead53 verified
Raw History Blame Contribute Delete
193 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.8038638322321172,
"eval_steps": 500,
"global_step": 11193,
"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": 0.8671253725414513,
"learning_rate": 2e-05,
"loss": 0.5332589149475098,
"step": 5090
},
{
"epoch": 0.366274059178397,
"grad_norm": 1.8518161862423437,
"learning_rate": 2e-05,
"loss": 0.5439261436462403,
"step": 5100
},
{
"epoch": 0.3669922436081586,
"grad_norm": 0.7392081443256451,
"learning_rate": 2e-05,
"loss": 0.5548827171325683,
"step": 5110
},
{
"epoch": 0.36771042803792015,
"grad_norm": 1.3189402012496558,
"learning_rate": 2e-05,
"loss": 0.5463480949401855,
"step": 5120
},
{
"epoch": 0.3684286124676817,
"grad_norm": 2.0924232204123596,
"learning_rate": 2e-05,
"loss": 0.5475778102874755,
"step": 5130
},
{
"epoch": 0.3691467968974433,
"grad_norm": 1.3798767942297876,
"learning_rate": 2e-05,
"loss": 0.5220169067382813,
"step": 5140
},
{
"epoch": 0.3698649813272048,
"grad_norm": 3.5370214674814715,
"learning_rate": 2e-05,
"loss": 0.5181848526000976,
"step": 5150
},
{
"epoch": 0.37058316575696637,
"grad_norm": 1.1742656367322117,
"learning_rate": 2e-05,
"loss": 0.5161781311035156,
"step": 5160
},
{
"epoch": 0.37130135018672794,
"grad_norm": 0.8051146242049202,
"learning_rate": 2e-05,
"loss": 0.5287814140319824,
"step": 5170
},
{
"epoch": 0.3720195346164895,
"grad_norm": 0.6983059164236506,
"learning_rate": 2e-05,
"loss": 0.5029990673065186,
"step": 5180
},
{
"epoch": 0.3727377190462511,
"grad_norm": 1.1890966558605038,
"learning_rate": 2e-05,
"loss": 0.5018293857574463,
"step": 5190
},
{
"epoch": 0.37345590347601265,
"grad_norm": 0.8176236211892003,
"learning_rate": 2e-05,
"loss": 0.520080280303955,
"step": 5200
},
{
"epoch": 0.3741740879057742,
"grad_norm": 0.7993900451498039,
"learning_rate": 2e-05,
"loss": 0.5241353511810303,
"step": 5210
},
{
"epoch": 0.3748922723355358,
"grad_norm": 0.733702206676448,
"learning_rate": 2e-05,
"loss": 0.5285841941833496,
"step": 5220
},
{
"epoch": 0.3756104567652973,
"grad_norm": 0.8255883781543882,
"learning_rate": 2e-05,
"loss": 0.532388162612915,
"step": 5230
},
{
"epoch": 0.3763286411950589,
"grad_norm": 1.039540488568116,
"learning_rate": 2e-05,
"loss": 0.5074212074279785,
"step": 5240
},
{
"epoch": 0.37704682562482045,
"grad_norm": 1.071380477631796,
"learning_rate": 2e-05,
"loss": 0.4883869171142578,
"step": 5250
},
{
"epoch": 0.377765010054582,
"grad_norm": 1.9185353655974138,
"learning_rate": 2e-05,
"loss": 0.5328169822692871,
"step": 5260
},
{
"epoch": 0.3784831944843436,
"grad_norm": 0.7853978937185732,
"learning_rate": 2e-05,
"loss": 0.5354257106781006,
"step": 5270
},
{
"epoch": 0.37920137891410516,
"grad_norm": 0.7134424246416942,
"learning_rate": 2e-05,
"loss": 0.5321937561035156,
"step": 5280
},
{
"epoch": 0.37991956334386673,
"grad_norm": 0.9011178633237298,
"learning_rate": 2e-05,
"loss": 0.5565654754638671,
"step": 5290
},
{
"epoch": 0.38063774777362824,
"grad_norm": 0.8286060017409398,
"learning_rate": 2e-05,
"loss": 0.536436653137207,
"step": 5300
},
{
"epoch": 0.3813559322033898,
"grad_norm": 0.9048514228440057,
"learning_rate": 2e-05,
"loss": 0.4888449192047119,
"step": 5310
},
{
"epoch": 0.3820741166331514,
"grad_norm": 2.5509663277307237,
"learning_rate": 2e-05,
"loss": 0.505039119720459,
"step": 5320
},
{
"epoch": 0.38279230106291295,
"grad_norm": 1.9528628519819875,
"learning_rate": 2e-05,
"loss": 0.5537528991699219,
"step": 5330
},
{
"epoch": 0.3835104854926745,
"grad_norm": 4.540438657847455,
"learning_rate": 2e-05,
"loss": 0.5073054313659668,
"step": 5340
},
{
"epoch": 0.3842286699224361,
"grad_norm": 1.7276858492967504,
"learning_rate": 2e-05,
"loss": 0.5421859741210937,
"step": 5350
},
{
"epoch": 0.38494685435219766,
"grad_norm": 0.9589188436330404,
"learning_rate": 2e-05,
"loss": 0.5715352058410644,
"step": 5360
},
{
"epoch": 0.3856650387819592,
"grad_norm": 0.8086342435573133,
"learning_rate": 2e-05,
"loss": 0.5158023834228516,
"step": 5370
},
{
"epoch": 0.38638322321172075,
"grad_norm": 0.8873607586806589,
"learning_rate": 2e-05,
"loss": 0.523341178894043,
"step": 5380
},
{
"epoch": 0.3871014076414823,
"grad_norm": 0.8484330319802704,
"learning_rate": 2e-05,
"loss": 0.5171842575073242,
"step": 5390
},
{
"epoch": 0.3878195920712439,
"grad_norm": 0.6717319465045217,
"learning_rate": 2e-05,
"loss": 0.5313539028167724,
"step": 5400
},
{
"epoch": 0.38853777650100546,
"grad_norm": 4.439744196657368,
"learning_rate": 2e-05,
"loss": 0.5450019836425781,
"step": 5410
},
{
"epoch": 0.38925596093076703,
"grad_norm": 1.8032293092463842,
"learning_rate": 2e-05,
"loss": 0.5292023181915283,
"step": 5420
},
{
"epoch": 0.3899741453605286,
"grad_norm": 1.1549168938141035,
"learning_rate": 2e-05,
"loss": 0.5448451519012452,
"step": 5430
},
{
"epoch": 0.39069232979029017,
"grad_norm": 1.557951912303909,
"learning_rate": 2e-05,
"loss": 0.504024887084961,
"step": 5440
},
{
"epoch": 0.3914105142200517,
"grad_norm": 0.8953994263800684,
"learning_rate": 2e-05,
"loss": 0.5561927795410156,
"step": 5450
},
{
"epoch": 0.39212869864981326,
"grad_norm": 0.776474289294985,
"learning_rate": 2e-05,
"loss": 0.5392631530761719,
"step": 5460
},
{
"epoch": 0.3928468830795748,
"grad_norm": 1.0056116720803987,
"learning_rate": 2e-05,
"loss": 0.5263529300689698,
"step": 5470
},
{
"epoch": 0.3935650675093364,
"grad_norm": 1.8688480844278623,
"learning_rate": 2e-05,
"loss": 0.5218052864074707,
"step": 5480
},
{
"epoch": 0.39428325193909797,
"grad_norm": 0.8105167392246526,
"learning_rate": 2e-05,
"loss": 0.5611165523529053,
"step": 5490
},
{
"epoch": 0.39500143636885954,
"grad_norm": 0.8084761426633206,
"learning_rate": 2e-05,
"loss": 0.5526279449462891,
"step": 5500
},
{
"epoch": 0.3957196207986211,
"grad_norm": 0.7700866718433312,
"learning_rate": 2e-05,
"loss": 0.538936996459961,
"step": 5510
},
{
"epoch": 0.3964378052283826,
"grad_norm": 1.0657431689557055,
"learning_rate": 2e-05,
"loss": 0.5506243705749512,
"step": 5520
},
{
"epoch": 0.3971559896581442,
"grad_norm": 0.8745275986637756,
"learning_rate": 2e-05,
"loss": 0.5275880813598632,
"step": 5530
},
{
"epoch": 0.39787417408790576,
"grad_norm": 0.9308238257341008,
"learning_rate": 2e-05,
"loss": 0.4923418045043945,
"step": 5540
},
{
"epoch": 0.39859235851766733,
"grad_norm": 0.766558679360876,
"learning_rate": 2e-05,
"loss": 0.5173656940460205,
"step": 5550
},
{
"epoch": 0.3993105429474289,
"grad_norm": 2.5226143307039846,
"learning_rate": 2e-05,
"loss": 0.5309979915618896,
"step": 5560
},
{
"epoch": 0.4000287273771905,
"grad_norm": 0.8794874574891549,
"learning_rate": 2e-05,
"loss": 0.5480818748474121,
"step": 5570
},
{
"epoch": 0.40074691180695204,
"grad_norm": 0.8080034444016956,
"learning_rate": 2e-05,
"loss": 0.5606730937957763,
"step": 5580
},
{
"epoch": 0.4014650962367136,
"grad_norm": 0.907140270351955,
"learning_rate": 2e-05,
"loss": 0.5224662303924561,
"step": 5590
},
{
"epoch": 0.40218328066647513,
"grad_norm": 1.1744743737870154,
"learning_rate": 2e-05,
"loss": 0.5013256072998047,
"step": 5600
},
{
"epoch": 0.4029014650962367,
"grad_norm": 0.9502763243618761,
"learning_rate": 2e-05,
"loss": 0.5188658714294434,
"step": 5610
},
{
"epoch": 0.40361964952599827,
"grad_norm": 0.836983749379159,
"learning_rate": 2e-05,
"loss": 0.5365310668945312,
"step": 5620
},
{
"epoch": 0.40433783395575984,
"grad_norm": 1.0909250419542207,
"learning_rate": 2e-05,
"loss": 0.525754165649414,
"step": 5630
},
{
"epoch": 0.4050560183855214,
"grad_norm": 0.9595120462140865,
"learning_rate": 2e-05,
"loss": 0.5007344722747803,
"step": 5640
},
{
"epoch": 0.405774202815283,
"grad_norm": 0.9884423585526875,
"learning_rate": 2e-05,
"loss": 0.5287074089050293,
"step": 5650
},
{
"epoch": 0.40649238724504455,
"grad_norm": 2.0139135573916533,
"learning_rate": 2e-05,
"loss": 0.5224942207336426,
"step": 5660
},
{
"epoch": 0.40721057167480607,
"grad_norm": 1.6077794416450646,
"learning_rate": 2e-05,
"loss": 0.5402818202972413,
"step": 5670
},
{
"epoch": 0.40792875610456764,
"grad_norm": 1.2846705363510933,
"learning_rate": 2e-05,
"loss": 0.5311152458190918,
"step": 5680
},
{
"epoch": 0.4086469405343292,
"grad_norm": 0.6547689504350582,
"learning_rate": 2e-05,
"loss": 0.5408687591552734,
"step": 5690
},
{
"epoch": 0.4093651249640908,
"grad_norm": 0.9403619444425615,
"learning_rate": 2e-05,
"loss": 0.534623098373413,
"step": 5700
},
{
"epoch": 0.41008330939385235,
"grad_norm": 1.8907345224993337,
"learning_rate": 2e-05,
"loss": 0.5487159729003906,
"step": 5710
},
{
"epoch": 0.4108014938236139,
"grad_norm": 0.8277579053282774,
"learning_rate": 2e-05,
"loss": 0.5145920753479004,
"step": 5720
},
{
"epoch": 0.4115196782533755,
"grad_norm": 0.8500504915667478,
"learning_rate": 2e-05,
"loss": 0.5577750682830811,
"step": 5730
},
{
"epoch": 0.412237862683137,
"grad_norm": 0.8766351752648883,
"learning_rate": 2e-05,
"loss": 0.5245782852172851,
"step": 5740
},
{
"epoch": 0.41295604711289857,
"grad_norm": 3.9356618171087514,
"learning_rate": 2e-05,
"loss": 0.5203507423400879,
"step": 5750
},
{
"epoch": 0.41367423154266014,
"grad_norm": 0.8394542728521265,
"learning_rate": 2e-05,
"loss": 0.5020275115966797,
"step": 5760
},
{
"epoch": 0.4143924159724217,
"grad_norm": 1.2301282234441133,
"learning_rate": 2e-05,
"loss": 0.5109591484069824,
"step": 5770
},
{
"epoch": 0.4151106004021833,
"grad_norm": 0.8540194330604614,
"learning_rate": 2e-05,
"loss": 0.5362766265869141,
"step": 5780
},
{
"epoch": 0.41582878483194485,
"grad_norm": 1.3115732158849696,
"learning_rate": 2e-05,
"loss": 0.5441376686096191,
"step": 5790
},
{
"epoch": 0.4165469692617064,
"grad_norm": 0.9891573038209267,
"learning_rate": 2e-05,
"loss": 0.5286791801452637,
"step": 5800
},
{
"epoch": 0.417265153691468,
"grad_norm": 1.0369349119087081,
"learning_rate": 2e-05,
"loss": 0.5240501880645752,
"step": 5810
},
{
"epoch": 0.4179833381212295,
"grad_norm": 0.6924684482192942,
"learning_rate": 2e-05,
"loss": 0.5057554244995117,
"step": 5820
},
{
"epoch": 0.4187015225509911,
"grad_norm": 1.1553489558444643,
"learning_rate": 2e-05,
"loss": 0.49388957023620605,
"step": 5830
},
{
"epoch": 0.41941970698075265,
"grad_norm": 0.9260613074570967,
"learning_rate": 2e-05,
"loss": 0.5467721939086914,
"step": 5840
},
{
"epoch": 0.4201378914105142,
"grad_norm": 0.8195187800487593,
"learning_rate": 2e-05,
"loss": 0.5344364166259765,
"step": 5850
},
{
"epoch": 0.4208560758402758,
"grad_norm": 27.366218903464574,
"learning_rate": 2e-05,
"loss": 0.5131311416625977,
"step": 5860
},
{
"epoch": 0.42157426027003736,
"grad_norm": 0.8103256215072813,
"learning_rate": 2e-05,
"loss": 0.5696263790130616,
"step": 5870
},
{
"epoch": 0.42229244469979893,
"grad_norm": 0.7260572666578229,
"learning_rate": 2e-05,
"loss": 0.5537211418151855,
"step": 5880
},
{
"epoch": 0.42301062912956044,
"grad_norm": 0.7455076690451402,
"learning_rate": 2e-05,
"loss": 0.5224467277526855,
"step": 5890
},
{
"epoch": 0.423728813559322,
"grad_norm": 0.8530894500971355,
"learning_rate": 2e-05,
"loss": 0.4967904567718506,
"step": 5900
},
{
"epoch": 0.4244469979890836,
"grad_norm": 0.7816275147165572,
"learning_rate": 2e-05,
"loss": 0.5174599647521972,
"step": 5910
},
{
"epoch": 0.42516518241884516,
"grad_norm": 0.9771385263047833,
"learning_rate": 2e-05,
"loss": 0.5489806175231934,
"step": 5920
},
{
"epoch": 0.4258833668486067,
"grad_norm": 2.240003677953534,
"learning_rate": 2e-05,
"loss": 0.5204599857330322,
"step": 5930
},
{
"epoch": 0.4266015512783683,
"grad_norm": 1.035549673371988,
"learning_rate": 2e-05,
"loss": 0.5292465209960937,
"step": 5940
},
{
"epoch": 0.42731973570812987,
"grad_norm": 1.4190268278299074,
"learning_rate": 2e-05,
"loss": 0.535629415512085,
"step": 5950
},
{
"epoch": 0.42803792013789144,
"grad_norm": 0.7458713966675397,
"learning_rate": 2e-05,
"loss": 0.5198601245880127,
"step": 5960
},
{
"epoch": 0.42875610456765295,
"grad_norm": 0.9547842725287439,
"learning_rate": 2e-05,
"loss": 0.5569463729858398,
"step": 5970
},
{
"epoch": 0.4294742889974145,
"grad_norm": 1.6072267797508677,
"learning_rate": 2e-05,
"loss": 0.5067478656768799,
"step": 5980
},
{
"epoch": 0.4301924734271761,
"grad_norm": 0.6249880704217256,
"learning_rate": 2e-05,
"loss": 0.5143586158752441,
"step": 5990
},
{
"epoch": 0.43091065785693766,
"grad_norm": 0.8291807899319975,
"learning_rate": 2e-05,
"loss": 0.5352517604827881,
"step": 6000
},
{
"epoch": 0.43162884228669923,
"grad_norm": 0.7796437047468323,
"learning_rate": 2e-05,
"loss": 0.5036755561828613,
"step": 6010
},
{
"epoch": 0.4323470267164608,
"grad_norm": 0.8008406423757182,
"learning_rate": 2e-05,
"loss": 0.5090443134307862,
"step": 6020
},
{
"epoch": 0.4330652111462224,
"grad_norm": 0.7694634166064175,
"learning_rate": 2e-05,
"loss": 0.5531556606292725,
"step": 6030
},
{
"epoch": 0.4337833955759839,
"grad_norm": 0.7420493720611807,
"learning_rate": 2e-05,
"loss": 0.5015220165252685,
"step": 6040
},
{
"epoch": 0.43450158000574546,
"grad_norm": 4.64584347221278,
"learning_rate": 2e-05,
"loss": 0.5164049148559571,
"step": 6050
},
{
"epoch": 0.43521976443550703,
"grad_norm": 0.9140495452925551,
"learning_rate": 2e-05,
"loss": 0.5333127021789551,
"step": 6060
},
{
"epoch": 0.4359379488652686,
"grad_norm": 1.863947303769377,
"learning_rate": 2e-05,
"loss": 0.5152500152587891,
"step": 6070
},
{
"epoch": 0.43665613329503017,
"grad_norm": 0.7468638697341918,
"learning_rate": 2e-05,
"loss": 0.5474298000335693,
"step": 6080
},
{
"epoch": 0.43737431772479174,
"grad_norm": 0.8701007622741522,
"learning_rate": 2e-05,
"loss": 0.5081675529479981,
"step": 6090
},
{
"epoch": 0.4380925021545533,
"grad_norm": 0.934465093526427,
"learning_rate": 2e-05,
"loss": 0.5175466060638427,
"step": 6100
},
{
"epoch": 0.4388106865843149,
"grad_norm": 0.8616975144200346,
"learning_rate": 2e-05,
"loss": 0.5210711479187011,
"step": 6110
},
{
"epoch": 0.4395288710140764,
"grad_norm": 0.7416920345061392,
"learning_rate": 2e-05,
"loss": 0.5216055393218995,
"step": 6120
},
{
"epoch": 0.44024705544383796,
"grad_norm": 1.1903720286358137,
"learning_rate": 2e-05,
"loss": 0.5488819122314453,
"step": 6130
},
{
"epoch": 0.44096523987359953,
"grad_norm": 1.6591977955683088,
"learning_rate": 2e-05,
"loss": 0.5320119380950927,
"step": 6140
},
{
"epoch": 0.4416834243033611,
"grad_norm": 1.163851828309418,
"learning_rate": 2e-05,
"loss": 0.5081817150115967,
"step": 6150
},
{
"epoch": 0.4424016087331227,
"grad_norm": 1.0848750719461473,
"learning_rate": 2e-05,
"loss": 0.5162517547607421,
"step": 6160
},
{
"epoch": 0.44311979316288425,
"grad_norm": 3.7839837678525874,
"learning_rate": 2e-05,
"loss": 0.5023531913757324,
"step": 6170
},
{
"epoch": 0.4438379775926458,
"grad_norm": 0.7885204732644386,
"learning_rate": 2e-05,
"loss": 0.5309329986572265,
"step": 6180
},
{
"epoch": 0.44455616202240733,
"grad_norm": 0.8617575017506026,
"learning_rate": 2e-05,
"loss": 0.5316021919250489,
"step": 6190
},
{
"epoch": 0.4452743464521689,
"grad_norm": 0.7891233799486785,
"learning_rate": 2e-05,
"loss": 0.5335870265960694,
"step": 6200
},
{
"epoch": 0.44599253088193047,
"grad_norm": 0.766826663453458,
"learning_rate": 2e-05,
"loss": 0.5129971027374267,
"step": 6210
},
{
"epoch": 0.44671071531169204,
"grad_norm": 1.0302501782204354,
"learning_rate": 2e-05,
"loss": 0.5231074333190918,
"step": 6220
},
{
"epoch": 0.4474288997414536,
"grad_norm": 0.6103086522733283,
"learning_rate": 2e-05,
"loss": 0.5257176876068115,
"step": 6230
},
{
"epoch": 0.4481470841712152,
"grad_norm": 0.6818759444109206,
"learning_rate": 2e-05,
"loss": 0.4979089260101318,
"step": 6240
},
{
"epoch": 0.44886526860097675,
"grad_norm": 0.7542308900696949,
"learning_rate": 2e-05,
"loss": 0.48744521141052244,
"step": 6250
},
{
"epoch": 0.44958345303073827,
"grad_norm": 0.914820519070103,
"learning_rate": 2e-05,
"loss": 0.5207819938659668,
"step": 6260
},
{
"epoch": 0.45030163746049984,
"grad_norm": 0.7633878018481595,
"learning_rate": 2e-05,
"loss": 0.5299099922180176,
"step": 6270
},
{
"epoch": 0.4510198218902614,
"grad_norm": 2.001935598999061,
"learning_rate": 2e-05,
"loss": 0.5330507278442382,
"step": 6280
},
{
"epoch": 0.451738006320023,
"grad_norm": 0.997774810084567,
"learning_rate": 2e-05,
"loss": 0.5312596797943115,
"step": 6290
},
{
"epoch": 0.45245619074978455,
"grad_norm": 1.0363442596735637,
"learning_rate": 2e-05,
"loss": 0.5274360179901123,
"step": 6300
},
{
"epoch": 0.4531743751795461,
"grad_norm": 0.6292653733771196,
"learning_rate": 2e-05,
"loss": 0.505595588684082,
"step": 6310
},
{
"epoch": 0.4538925596093077,
"grad_norm": 0.9004258652312962,
"learning_rate": 2e-05,
"loss": 0.534788703918457,
"step": 6320
},
{
"epoch": 0.45461074403906926,
"grad_norm": 0.7616652397171051,
"learning_rate": 2e-05,
"loss": 0.5294042587280273,
"step": 6330
},
{
"epoch": 0.4553289284688308,
"grad_norm": 0.7213294769887223,
"learning_rate": 2e-05,
"loss": 0.5184205055236817,
"step": 6340
},
{
"epoch": 0.45604711289859234,
"grad_norm": 1.0071587089572807,
"learning_rate": 2e-05,
"loss": 0.5413650512695313,
"step": 6350
},
{
"epoch": 0.4567652973283539,
"grad_norm": 1.0292994129605377,
"learning_rate": 2e-05,
"loss": 0.5207261562347412,
"step": 6360
},
{
"epoch": 0.4574834817581155,
"grad_norm": 0.6990183168538171,
"learning_rate": 2e-05,
"loss": 0.5392344951629638,
"step": 6370
},
{
"epoch": 0.45820166618787705,
"grad_norm": 0.972350204087938,
"learning_rate": 2e-05,
"loss": 0.4889195442199707,
"step": 6380
},
{
"epoch": 0.4589198506176386,
"grad_norm": 0.7779311371843183,
"learning_rate": 2e-05,
"loss": 0.47156562805175783,
"step": 6390
},
{
"epoch": 0.4596380350474002,
"grad_norm": 0.8911321138522645,
"learning_rate": 2e-05,
"loss": 0.4952733039855957,
"step": 6400
},
{
"epoch": 0.4603562194771617,
"grad_norm": 3.6582198721248034,
"learning_rate": 2e-05,
"loss": 0.5478610515594482,
"step": 6410
},
{
"epoch": 0.4610744039069233,
"grad_norm": 2.1146528228991355,
"learning_rate": 2e-05,
"loss": 0.5329720497131347,
"step": 6420
},
{
"epoch": 0.46179258833668485,
"grad_norm": 0.6594625834230033,
"learning_rate": 2e-05,
"loss": 0.5087614059448242,
"step": 6430
},
{
"epoch": 0.4625107727664464,
"grad_norm": 0.8925611843063261,
"learning_rate": 2e-05,
"loss": 0.5519828796386719,
"step": 6440
},
{
"epoch": 0.463228957196208,
"grad_norm": 0.747142126469513,
"learning_rate": 2e-05,
"loss": 0.5004665374755859,
"step": 6450
},
{
"epoch": 0.46394714162596956,
"grad_norm": 0.7411024127252138,
"learning_rate": 2e-05,
"loss": 0.512296485900879,
"step": 6460
},
{
"epoch": 0.46466532605573113,
"grad_norm": 0.7078827195428221,
"learning_rate": 2e-05,
"loss": 0.5357799053192138,
"step": 6470
},
{
"epoch": 0.4653835104854927,
"grad_norm": 0.839454498365693,
"learning_rate": 2e-05,
"loss": 0.5336461544036866,
"step": 6480
},
{
"epoch": 0.4661016949152542,
"grad_norm": 1.3588384065949128,
"learning_rate": 2e-05,
"loss": 0.5107567787170411,
"step": 6490
},
{
"epoch": 0.4668198793450158,
"grad_norm": 1.3852997105008118,
"learning_rate": 2e-05,
"loss": 0.51701340675354,
"step": 6500
},
{
"epoch": 0.46753806377477736,
"grad_norm": 1.4073176002109269,
"learning_rate": 2e-05,
"loss": 0.48174543380737306,
"step": 6510
},
{
"epoch": 0.4682562482045389,
"grad_norm": 0.8924861265149928,
"learning_rate": 2e-05,
"loss": 0.5137310981750488,
"step": 6520
},
{
"epoch": 0.4689744326343005,
"grad_norm": 0.7737072490931616,
"learning_rate": 2e-05,
"loss": 0.5131258487701416,
"step": 6530
},
{
"epoch": 0.46969261706406207,
"grad_norm": 0.9696773985808531,
"learning_rate": 2e-05,
"loss": 0.5347423553466797,
"step": 6540
},
{
"epoch": 0.47041080149382364,
"grad_norm": 0.7704459739424128,
"learning_rate": 2e-05,
"loss": 0.5029036998748779,
"step": 6550
},
{
"epoch": 0.47112898592358515,
"grad_norm": 0.6971145027730525,
"learning_rate": 2e-05,
"loss": 0.5131199836730957,
"step": 6560
},
{
"epoch": 0.4718471703533467,
"grad_norm": 0.7013430432071835,
"learning_rate": 2e-05,
"loss": 0.5082266330718994,
"step": 6570
},
{
"epoch": 0.4725653547831083,
"grad_norm": 0.6895514264818592,
"learning_rate": 2e-05,
"loss": 0.5182147026062012,
"step": 6580
},
{
"epoch": 0.47328353921286986,
"grad_norm": 0.6092525675578216,
"learning_rate": 2e-05,
"loss": 0.5235995292663574,
"step": 6590
},
{
"epoch": 0.47400172364263143,
"grad_norm": 0.7127170661262854,
"learning_rate": 2e-05,
"loss": 0.49425926208496096,
"step": 6600
},
{
"epoch": 0.474719908072393,
"grad_norm": 0.8484018530852713,
"learning_rate": 2e-05,
"loss": 0.510321044921875,
"step": 6610
},
{
"epoch": 0.4754380925021546,
"grad_norm": 0.714534413474948,
"learning_rate": 2e-05,
"loss": 0.5539217948913574,
"step": 6620
},
{
"epoch": 0.4761562769319161,
"grad_norm": 0.8732638802378228,
"learning_rate": 2e-05,
"loss": 0.5318286895751954,
"step": 6630
},
{
"epoch": 0.47687446136167766,
"grad_norm": 1.2323092918034892,
"learning_rate": 2e-05,
"loss": 0.5199755191802978,
"step": 6640
},
{
"epoch": 0.47759264579143923,
"grad_norm": 0.7414208671728494,
"learning_rate": 2e-05,
"loss": 0.5017186164855957,
"step": 6650
},
{
"epoch": 0.4783108302212008,
"grad_norm": 0.9091138295377136,
"learning_rate": 2e-05,
"loss": 0.5028333187103271,
"step": 6660
},
{
"epoch": 0.47902901465096237,
"grad_norm": 1.0229512635570075,
"learning_rate": 2e-05,
"loss": 0.5375429153442383,
"step": 6670
},
{
"epoch": 0.47974719908072394,
"grad_norm": 1.0258530094113092,
"learning_rate": 2e-05,
"loss": 0.5018298149108886,
"step": 6680
},
{
"epoch": 0.4804653835104855,
"grad_norm": 0.700829383483421,
"learning_rate": 2e-05,
"loss": 0.5151088237762451,
"step": 6690
},
{
"epoch": 0.4811835679402471,
"grad_norm": 0.8152546270298604,
"learning_rate": 2e-05,
"loss": 0.5204953670501709,
"step": 6700
},
{
"epoch": 0.4819017523700086,
"grad_norm": 0.8542602502767465,
"learning_rate": 2e-05,
"loss": 0.5167047500610351,
"step": 6710
},
{
"epoch": 0.48261993679977017,
"grad_norm": 0.8992754223541527,
"learning_rate": 2e-05,
"loss": 0.5057715892791748,
"step": 6720
},
{
"epoch": 0.48333812122953174,
"grad_norm": 0.7199038026164994,
"learning_rate": 2e-05,
"loss": 0.5120359897613526,
"step": 6730
},
{
"epoch": 0.4840563056592933,
"grad_norm": 1.2288205967469437,
"learning_rate": 2e-05,
"loss": 0.5329171657562256,
"step": 6740
},
{
"epoch": 0.4847744900890549,
"grad_norm": 0.8404746888839109,
"learning_rate": 2e-05,
"loss": 0.5213992595672607,
"step": 6750
},
{
"epoch": 0.48549267451881645,
"grad_norm": 1.1835417298129078,
"learning_rate": 2e-05,
"loss": 0.5152032375335693,
"step": 6760
},
{
"epoch": 0.486210858948578,
"grad_norm": 0.9615234318450225,
"learning_rate": 2e-05,
"loss": 0.5006381034851074,
"step": 6770
},
{
"epoch": 0.48692904337833953,
"grad_norm": 1.329668077951727,
"learning_rate": 2e-05,
"loss": 0.5205175399780273,
"step": 6780
},
{
"epoch": 0.4876472278081011,
"grad_norm": 0.6443294107715632,
"learning_rate": 2e-05,
"loss": 0.5255660057067871,
"step": 6790
},
{
"epoch": 0.4883654122378627,
"grad_norm": 0.803996743157343,
"learning_rate": 2e-05,
"loss": 0.5074993133544922,
"step": 6800
},
{
"epoch": 0.48908359666762424,
"grad_norm": 0.7588310938434683,
"learning_rate": 2e-05,
"loss": 0.5150827407836914,
"step": 6810
},
{
"epoch": 0.4898017810973858,
"grad_norm": 1.0364989573417396,
"learning_rate": 2e-05,
"loss": 0.4786884307861328,
"step": 6820
},
{
"epoch": 0.4905199655271474,
"grad_norm": 1.4584468663774266,
"learning_rate": 2e-05,
"loss": 0.5116088390350342,
"step": 6830
},
{
"epoch": 0.49123814995690895,
"grad_norm": 1.0426184531320342,
"learning_rate": 2e-05,
"loss": 0.50509672164917,
"step": 6840
},
{
"epoch": 0.4919563343866705,
"grad_norm": 0.7676452395513347,
"learning_rate": 2e-05,
"loss": 0.5092341423034668,
"step": 6850
},
{
"epoch": 0.49267451881643204,
"grad_norm": 0.8276951020934097,
"learning_rate": 2e-05,
"loss": 0.491436767578125,
"step": 6860
},
{
"epoch": 0.4933927032461936,
"grad_norm": 0.6947177209793752,
"learning_rate": 2e-05,
"loss": 0.5210543632507324,
"step": 6870
},
{
"epoch": 0.4941108876759552,
"grad_norm": 0.9400527891534112,
"learning_rate": 2e-05,
"loss": 0.513559627532959,
"step": 6880
},
{
"epoch": 0.49482907210571675,
"grad_norm": 0.741115171613497,
"learning_rate": 2e-05,
"loss": 0.5199744701385498,
"step": 6890
},
{
"epoch": 0.4955472565354783,
"grad_norm": 1.3619965727261578,
"learning_rate": 2e-05,
"loss": 0.5108253002166748,
"step": 6900
},
{
"epoch": 0.4962654409652399,
"grad_norm": 0.9623508341472918,
"learning_rate": 2e-05,
"loss": 0.5400260925292969,
"step": 6910
},
{
"epoch": 0.49698362539500146,
"grad_norm": 0.9787251587070064,
"learning_rate": 2e-05,
"loss": 0.5064085960388184,
"step": 6920
},
{
"epoch": 0.497701809824763,
"grad_norm": 1.081125482137681,
"learning_rate": 2e-05,
"loss": 0.5072525978088379,
"step": 6930
},
{
"epoch": 0.49841999425452455,
"grad_norm": 1.3313792083198392,
"learning_rate": 2e-05,
"loss": 0.482879638671875,
"step": 6940
},
{
"epoch": 0.4991381786842861,
"grad_norm": 0.93513024204936,
"learning_rate": 2e-05,
"loss": 0.5279765605926514,
"step": 6950
},
{
"epoch": 0.4998563631140477,
"grad_norm": 0.7524692824637937,
"learning_rate": 2e-05,
"loss": 0.5047333717346192,
"step": 6960
},
{
"epoch": 0.5005745475438093,
"grad_norm": 0.915294034974198,
"learning_rate": 2e-05,
"loss": 0.521539306640625,
"step": 6970
},
{
"epoch": 0.5012927319735708,
"grad_norm": 1.014629372473378,
"learning_rate": 2e-05,
"loss": 0.532783317565918,
"step": 6980
},
{
"epoch": 0.5020109164033324,
"grad_norm": 1.2194440273670863,
"learning_rate": 2e-05,
"loss": 0.5128929138183593,
"step": 6990
},
{
"epoch": 0.502729100833094,
"grad_norm": 1.0596730186081031,
"learning_rate": 2e-05,
"loss": 0.5279033184051514,
"step": 7000
},
{
"epoch": 0.5034472852628555,
"grad_norm": 0.8083690357228798,
"learning_rate": 2e-05,
"loss": 0.5104133605957031,
"step": 7010
},
{
"epoch": 0.5041654696926171,
"grad_norm": 0.6801443540998167,
"learning_rate": 2e-05,
"loss": 0.527102279663086,
"step": 7020
},
{
"epoch": 0.5048836541223787,
"grad_norm": 0.7756206046576168,
"learning_rate": 2e-05,
"loss": 0.4832464218139648,
"step": 7030
},
{
"epoch": 0.5056018385521401,
"grad_norm": 0.9019380579554936,
"learning_rate": 2e-05,
"loss": 0.4857687473297119,
"step": 7040
},
{
"epoch": 0.5063200229819017,
"grad_norm": 0.9090375983532191,
"learning_rate": 2e-05,
"loss": 0.5374709606170655,
"step": 7050
},
{
"epoch": 0.5070382074116633,
"grad_norm": 0.8858279569923521,
"learning_rate": 2e-05,
"loss": 0.513466215133667,
"step": 7060
},
{
"epoch": 0.5077563918414248,
"grad_norm": 1.526429770908195,
"learning_rate": 2e-05,
"loss": 0.5261812210083008,
"step": 7070
},
{
"epoch": 0.5084745762711864,
"grad_norm": 0.7607242619908129,
"learning_rate": 2e-05,
"loss": 0.5106912136077881,
"step": 7080
},
{
"epoch": 0.509192760700948,
"grad_norm": 2.916806745740047,
"learning_rate": 2e-05,
"loss": 0.5033746719360351,
"step": 7090
},
{
"epoch": 0.5099109451307096,
"grad_norm": 0.6917777993364783,
"learning_rate": 2e-05,
"loss": 0.5160577774047852,
"step": 7100
},
{
"epoch": 0.5106291295604711,
"grad_norm": 0.830373330248796,
"learning_rate": 2e-05,
"loss": 0.5160553932189942,
"step": 7110
},
{
"epoch": 0.5113473139902327,
"grad_norm": 0.9577880876742136,
"learning_rate": 2e-05,
"loss": 0.5296999931335449,
"step": 7120
},
{
"epoch": 0.5120654984199943,
"grad_norm": 1.524221246488792,
"learning_rate": 2e-05,
"loss": 0.5396167755126953,
"step": 7130
},
{
"epoch": 0.5127836828497558,
"grad_norm": 1.0091619152135625,
"learning_rate": 2e-05,
"loss": 0.4928737640380859,
"step": 7140
},
{
"epoch": 0.5135018672795174,
"grad_norm": 2.565825640971298,
"learning_rate": 2e-05,
"loss": 0.49898929595947267,
"step": 7150
},
{
"epoch": 0.514220051709279,
"grad_norm": 0.9100408239623469,
"learning_rate": 2e-05,
"loss": 0.5215423107147217,
"step": 7160
},
{
"epoch": 0.5149382361390406,
"grad_norm": 0.815804987302966,
"learning_rate": 2e-05,
"loss": 0.5159441471099854,
"step": 7170
},
{
"epoch": 0.5156564205688021,
"grad_norm": 3.494960314307889,
"learning_rate": 2e-05,
"loss": 0.4860187530517578,
"step": 7180
},
{
"epoch": 0.5163746049985636,
"grad_norm": 1.2507875003414959,
"learning_rate": 2e-05,
"loss": 0.5292512893676757,
"step": 7190
},
{
"epoch": 0.5170927894283252,
"grad_norm": 0.915771895448318,
"learning_rate": 2e-05,
"loss": 0.5130393981933594,
"step": 7200
},
{
"epoch": 0.5178109738580867,
"grad_norm": 1.1879129825285109,
"learning_rate": 2e-05,
"loss": 0.5084530353546143,
"step": 7210
},
{
"epoch": 0.5185291582878483,
"grad_norm": 0.9405592075187831,
"learning_rate": 2e-05,
"loss": 0.5063467502593995,
"step": 7220
},
{
"epoch": 0.5192473427176099,
"grad_norm": 1.140126982815008,
"learning_rate": 2e-05,
"loss": 0.5135111331939697,
"step": 7230
},
{
"epoch": 0.5199655271473714,
"grad_norm": 1.0838648410517613,
"learning_rate": 2e-05,
"loss": 0.5293627262115479,
"step": 7240
},
{
"epoch": 0.520683711577133,
"grad_norm": 0.7311446871161834,
"learning_rate": 2e-05,
"loss": 0.5027955532073974,
"step": 7250
},
{
"epoch": 0.5214018960068946,
"grad_norm": 0.9291644390753973,
"learning_rate": 2e-05,
"loss": 0.5069394588470459,
"step": 7260
},
{
"epoch": 0.5221200804366561,
"grad_norm": 1.123260487877758,
"learning_rate": 2e-05,
"loss": 0.5379010200500488,
"step": 7270
},
{
"epoch": 0.5228382648664177,
"grad_norm": 1.0347219662718548,
"learning_rate": 2e-05,
"loss": 0.4997987747192383,
"step": 7280
},
{
"epoch": 0.5235564492961793,
"grad_norm": 1.2672590317341805,
"learning_rate": 2e-05,
"loss": 0.5212868690490723,
"step": 7290
},
{
"epoch": 0.5242746337259409,
"grad_norm": 1.5621975424296257,
"learning_rate": 2e-05,
"loss": 0.5097826004028321,
"step": 7300
},
{
"epoch": 0.5249928181557024,
"grad_norm": 0.7465240452166718,
"learning_rate": 2e-05,
"loss": 0.5357519149780273,
"step": 7310
},
{
"epoch": 0.525711002585464,
"grad_norm": 1.5224198677712948,
"learning_rate": 2e-05,
"loss": 0.5308655261993408,
"step": 7320
},
{
"epoch": 0.5264291870152256,
"grad_norm": 2.026105732375691,
"learning_rate": 2e-05,
"loss": 0.5084532737731934,
"step": 7330
},
{
"epoch": 0.527147371444987,
"grad_norm": 1.348927263221628,
"learning_rate": 2e-05,
"loss": 0.52230863571167,
"step": 7340
},
{
"epoch": 0.5278655558747486,
"grad_norm": 2.6636859768113648,
"learning_rate": 2e-05,
"loss": 0.5366877555847168,
"step": 7350
},
{
"epoch": 0.5285837403045102,
"grad_norm": 0.8618117773621916,
"learning_rate": 2e-05,
"loss": 0.5040595054626464,
"step": 7360
},
{
"epoch": 0.5293019247342717,
"grad_norm": 0.8458990891464304,
"learning_rate": 2e-05,
"loss": 0.5055989265441895,
"step": 7370
},
{
"epoch": 0.5300201091640333,
"grad_norm": 0.9956864721970093,
"learning_rate": 2e-05,
"loss": 0.5504654884338379,
"step": 7380
},
{
"epoch": 0.5307382935937949,
"grad_norm": 1.0064317906818125,
"learning_rate": 2e-05,
"loss": 0.5028414249420166,
"step": 7390
},
{
"epoch": 0.5314564780235564,
"grad_norm": 1.335574503381604,
"learning_rate": 2e-05,
"loss": 0.503629493713379,
"step": 7400
},
{
"epoch": 0.532174662453318,
"grad_norm": 0.9304180576843245,
"learning_rate": 2e-05,
"loss": 0.5050220489501953,
"step": 7410
},
{
"epoch": 0.5328928468830796,
"grad_norm": 2.435869791591822,
"learning_rate": 2e-05,
"loss": 0.4897305488586426,
"step": 7420
},
{
"epoch": 0.5336110313128412,
"grad_norm": 0.8929208107847408,
"learning_rate": 2e-05,
"loss": 0.5367134094238282,
"step": 7430
},
{
"epoch": 0.5343292157426027,
"grad_norm": 5.964530253739263,
"learning_rate": 2e-05,
"loss": 0.5006175994873047,
"step": 7440
},
{
"epoch": 0.5350474001723643,
"grad_norm": 1.8459409185857834,
"learning_rate": 2e-05,
"loss": 0.5164065361022949,
"step": 7450
},
{
"epoch": 0.5357655846021259,
"grad_norm": 2.468176367646822,
"learning_rate": 2e-05,
"loss": 0.529288387298584,
"step": 7460
},
{
"epoch": 0.5364837690318874,
"grad_norm": 0.8001616641858472,
"learning_rate": 2e-05,
"loss": 0.4970420837402344,
"step": 7470
},
{
"epoch": 0.537201953461649,
"grad_norm": 0.8993396236957818,
"learning_rate": 2e-05,
"loss": 0.5046211242675781,
"step": 7480
},
{
"epoch": 0.5379201378914105,
"grad_norm": 3.2243003419580476,
"learning_rate": 2e-05,
"loss": 0.5483418941497803,
"step": 7490
},
{
"epoch": 0.538638322321172,
"grad_norm": 0.7258029399794491,
"learning_rate": 2e-05,
"loss": 0.49911131858825686,
"step": 7500
},
{
"epoch": 0.5393565067509336,
"grad_norm": 0.7956516445380408,
"learning_rate": 2e-05,
"loss": 0.4908426761627197,
"step": 7510
},
{
"epoch": 0.5400746911806952,
"grad_norm": 1.0973276287842086,
"learning_rate": 2e-05,
"loss": 0.4936198711395264,
"step": 7520
},
{
"epoch": 0.5407928756104567,
"grad_norm": 0.7876930414255386,
"learning_rate": 2e-05,
"loss": 0.5183720111846923,
"step": 7530
},
{
"epoch": 0.5415110600402183,
"grad_norm": 0.9965685375501749,
"learning_rate": 2e-05,
"loss": 0.4960949897766113,
"step": 7540
},
{
"epoch": 0.5422292444699799,
"grad_norm": 1.0692615816007238,
"learning_rate": 2e-05,
"loss": 0.5305309295654297,
"step": 7550
},
{
"epoch": 0.5429474288997415,
"grad_norm": 0.7647171205315241,
"learning_rate": 2e-05,
"loss": 0.4955049991607666,
"step": 7560
},
{
"epoch": 0.543665613329503,
"grad_norm": 2.5055618835637574,
"learning_rate": 2e-05,
"loss": 0.47725300788879393,
"step": 7570
},
{
"epoch": 0.5443837977592646,
"grad_norm": 0.8092532781298774,
"learning_rate": 2e-05,
"loss": 0.5179968357086182,
"step": 7580
},
{
"epoch": 0.5451019821890262,
"grad_norm": 0.9591463041255972,
"learning_rate": 2e-05,
"loss": 0.5149166584014893,
"step": 7590
},
{
"epoch": 0.5458201666187877,
"grad_norm": 1.7958418206751419,
"learning_rate": 2e-05,
"loss": 0.5320939540863037,
"step": 7600
},
{
"epoch": 0.5465383510485493,
"grad_norm": 1.870397044677356,
"learning_rate": 2e-05,
"loss": 0.5256160736083985,
"step": 7610
},
{
"epoch": 0.5472565354783109,
"grad_norm": 1.0935245054549978,
"learning_rate": 2e-05,
"loss": 0.5159531116485596,
"step": 7620
},
{
"epoch": 0.5479747199080723,
"grad_norm": 0.6744932185130895,
"learning_rate": 2e-05,
"loss": 0.5272167205810547,
"step": 7630
},
{
"epoch": 0.5486929043378339,
"grad_norm": 0.9336853985371281,
"learning_rate": 2e-05,
"loss": 0.5258305072784424,
"step": 7640
},
{
"epoch": 0.5494110887675955,
"grad_norm": 0.8858754200338049,
"learning_rate": 2e-05,
"loss": 0.5083936214447021,
"step": 7650
},
{
"epoch": 0.550129273197357,
"grad_norm": 1.0864112527251981,
"learning_rate": 2e-05,
"loss": 0.5313797950744629,
"step": 7660
},
{
"epoch": 0.5508474576271186,
"grad_norm": 1.1845584365212702,
"learning_rate": 2e-05,
"loss": 0.5332317352294922,
"step": 7670
},
{
"epoch": 0.5515656420568802,
"grad_norm": 0.5822329774373377,
"learning_rate": 2e-05,
"loss": 0.5068902492523193,
"step": 7680
},
{
"epoch": 0.5522838264866418,
"grad_norm": 1.4850466911837648,
"learning_rate": 2e-05,
"loss": 0.5174357414245605,
"step": 7690
},
{
"epoch": 0.5530020109164033,
"grad_norm": 0.834434151124856,
"learning_rate": 2e-05,
"loss": 0.519992446899414,
"step": 7700
},
{
"epoch": 0.5537201953461649,
"grad_norm": 0.771487641523,
"learning_rate": 2e-05,
"loss": 0.5236385345458985,
"step": 7710
},
{
"epoch": 0.5544383797759265,
"grad_norm": 1.1943889324445582,
"learning_rate": 2e-05,
"loss": 0.5304265022277832,
"step": 7720
},
{
"epoch": 0.555156564205688,
"grad_norm": 0.7851518168635505,
"learning_rate": 2e-05,
"loss": 0.512861967086792,
"step": 7730
},
{
"epoch": 0.5558747486354496,
"grad_norm": 2.2101493906531524,
"learning_rate": 2e-05,
"loss": 0.5239198207855225,
"step": 7740
},
{
"epoch": 0.5565929330652112,
"grad_norm": 1.1967768525252627,
"learning_rate": 2e-05,
"loss": 0.536719799041748,
"step": 7750
},
{
"epoch": 0.5573111174949728,
"grad_norm": 0.8906351829337809,
"learning_rate": 2e-05,
"loss": 0.5058685302734375,
"step": 7760
},
{
"epoch": 0.5580293019247343,
"grad_norm": 0.8693073416497872,
"learning_rate": 2e-05,
"loss": 0.5323287963867187,
"step": 7770
},
{
"epoch": 0.5587474863544958,
"grad_norm": 0.7356601422485951,
"learning_rate": 2e-05,
"loss": 0.5065886497497558,
"step": 7780
},
{
"epoch": 0.5594656707842574,
"grad_norm": 0.6687923277377099,
"learning_rate": 2e-05,
"loss": 0.5112338066101074,
"step": 7790
},
{
"epoch": 0.5601838552140189,
"grad_norm": 0.7113557084365317,
"learning_rate": 2e-05,
"loss": 0.47762560844421387,
"step": 7800
},
{
"epoch": 0.5609020396437805,
"grad_norm": 0.8141012612329843,
"learning_rate": 2e-05,
"loss": 0.4991456031799316,
"step": 7810
},
{
"epoch": 0.5616202240735421,
"grad_norm": 0.787120338368763,
"learning_rate": 2e-05,
"loss": 0.5185105800628662,
"step": 7820
},
{
"epoch": 0.5623384085033036,
"grad_norm": 1.060427741807932,
"learning_rate": 2e-05,
"loss": 0.5031709671020508,
"step": 7830
},
{
"epoch": 0.5630565929330652,
"grad_norm": 0.7448093906054303,
"learning_rate": 2e-05,
"loss": 0.472198486328125,
"step": 7840
},
{
"epoch": 0.5637747773628268,
"grad_norm": 0.899506993834914,
"learning_rate": 2e-05,
"loss": 0.51429443359375,
"step": 7850
},
{
"epoch": 0.5644929617925883,
"grad_norm": 0.736011302968795,
"learning_rate": 2e-05,
"loss": 0.4995328426361084,
"step": 7860
},
{
"epoch": 0.5652111462223499,
"grad_norm": 0.746283848886023,
"learning_rate": 2e-05,
"loss": 0.5002707004547119,
"step": 7870
},
{
"epoch": 0.5659293306521115,
"grad_norm": 0.8402744094454515,
"learning_rate": 2e-05,
"loss": 0.5101712703704834,
"step": 7880
},
{
"epoch": 0.566647515081873,
"grad_norm": 0.7052511684135355,
"learning_rate": 2e-05,
"loss": 0.4966568946838379,
"step": 7890
},
{
"epoch": 0.5673656995116346,
"grad_norm": 0.7951822228215137,
"learning_rate": 2e-05,
"loss": 0.5166773796081543,
"step": 7900
},
{
"epoch": 0.5680838839413962,
"grad_norm": 0.8650455566916687,
"learning_rate": 2e-05,
"loss": 0.5096234321594239,
"step": 7910
},
{
"epoch": 0.5688020683711578,
"grad_norm": 0.6848642348671272,
"learning_rate": 2e-05,
"loss": 0.46970043182373045,
"step": 7920
},
{
"epoch": 0.5695202528009192,
"grad_norm": 0.6157571883908557,
"learning_rate": 2e-05,
"loss": 0.5021026611328125,
"step": 7930
},
{
"epoch": 0.5702384372306808,
"grad_norm": 0.6831186053495871,
"learning_rate": 2e-05,
"loss": 0.5141005992889405,
"step": 7940
},
{
"epoch": 0.5709566216604424,
"grad_norm": 0.7564939431221042,
"learning_rate": 2e-05,
"loss": 0.5459128856658936,
"step": 7950
},
{
"epoch": 0.5716748060902039,
"grad_norm": 0.7022768522025966,
"learning_rate": 2e-05,
"loss": 0.5102225780487061,
"step": 7960
},
{
"epoch": 0.5723929905199655,
"grad_norm": 0.6577457603860399,
"learning_rate": 2e-05,
"loss": 0.5143271446228027,
"step": 7970
},
{
"epoch": 0.5731111749497271,
"grad_norm": 0.888719396205061,
"learning_rate": 2e-05,
"loss": 0.4966334342956543,
"step": 7980
},
{
"epoch": 0.5738293593794886,
"grad_norm": 0.9303975380147393,
"learning_rate": 2e-05,
"loss": 0.5003435611724854,
"step": 7990
},
{
"epoch": 0.5745475438092502,
"grad_norm": 0.6663634670198717,
"learning_rate": 2e-05,
"loss": 0.49073519706726076,
"step": 8000
},
{
"epoch": 0.5752657282390118,
"grad_norm": 0.7392377131432697,
"learning_rate": 2e-05,
"loss": 0.5131159782409668,
"step": 8010
},
{
"epoch": 0.5759839126687734,
"grad_norm": 1.5728492147654642,
"learning_rate": 2e-05,
"loss": 0.5022923469543457,
"step": 8020
},
{
"epoch": 0.5767020970985349,
"grad_norm": 0.6344757102162765,
"learning_rate": 2e-05,
"loss": 0.4947488307952881,
"step": 8030
},
{
"epoch": 0.5774202815282965,
"grad_norm": 0.6662399243023269,
"learning_rate": 2e-05,
"loss": 0.5053641319274902,
"step": 8040
},
{
"epoch": 0.5781384659580581,
"grad_norm": 0.7851144646965664,
"learning_rate": 2e-05,
"loss": 0.5437193870544433,
"step": 8050
},
{
"epoch": 0.5788566503878196,
"grad_norm": 0.7737883906100125,
"learning_rate": 2e-05,
"loss": 0.4894735336303711,
"step": 8060
},
{
"epoch": 0.5795748348175812,
"grad_norm": 0.8322381729647296,
"learning_rate": 2e-05,
"loss": 0.5251737117767334,
"step": 8070
},
{
"epoch": 0.5802930192473427,
"grad_norm": 0.6166783511381956,
"learning_rate": 2e-05,
"loss": 0.513679027557373,
"step": 8080
},
{
"epoch": 0.5810112036771042,
"grad_norm": 0.70638704262371,
"learning_rate": 2e-05,
"loss": 0.532796049118042,
"step": 8090
},
{
"epoch": 0.5817293881068658,
"grad_norm": 0.9562141982943347,
"learning_rate": 2e-05,
"loss": 0.4967831611633301,
"step": 8100
},
{
"epoch": 0.5824475725366274,
"grad_norm": 0.8308460244387134,
"learning_rate": 2e-05,
"loss": 0.49094562530517577,
"step": 8110
},
{
"epoch": 0.583165756966389,
"grad_norm": 0.8546483459910912,
"learning_rate": 2e-05,
"loss": 0.5447869300842285,
"step": 8120
},
{
"epoch": 0.5838839413961505,
"grad_norm": 0.9296760210227759,
"learning_rate": 2e-05,
"loss": 0.4919291973114014,
"step": 8130
},
{
"epoch": 0.5846021258259121,
"grad_norm": 0.6938004998480967,
"learning_rate": 2e-05,
"loss": 0.5002818584442139,
"step": 8140
},
{
"epoch": 0.5853203102556737,
"grad_norm": 0.7098215376735001,
"learning_rate": 2e-05,
"loss": 0.4962301254272461,
"step": 8150
},
{
"epoch": 0.5860384946854352,
"grad_norm": 0.8462055529968932,
"learning_rate": 2e-05,
"loss": 0.5116305828094483,
"step": 8160
},
{
"epoch": 0.5867566791151968,
"grad_norm": 0.8252569182079765,
"learning_rate": 2e-05,
"loss": 0.5037623405456543,
"step": 8170
},
{
"epoch": 0.5874748635449584,
"grad_norm": 2.011160086261593,
"learning_rate": 2e-05,
"loss": 0.5188362121582031,
"step": 8180
},
{
"epoch": 0.5881930479747199,
"grad_norm": 0.9252415993521379,
"learning_rate": 2e-05,
"loss": 0.49549198150634766,
"step": 8190
},
{
"epoch": 0.5889112324044815,
"grad_norm": 0.9972448561942079,
"learning_rate": 2e-05,
"loss": 0.530819320678711,
"step": 8200
},
{
"epoch": 0.5896294168342431,
"grad_norm": 0.7707956147627286,
"learning_rate": 2e-05,
"loss": 0.5205513000488281,
"step": 8210
},
{
"epoch": 0.5903476012640047,
"grad_norm": 1.757948346241209,
"learning_rate": 2e-05,
"loss": 0.5204066276550293,
"step": 8220
},
{
"epoch": 0.5910657856937661,
"grad_norm": 0.6973748867763332,
"learning_rate": 2e-05,
"loss": 0.500050163269043,
"step": 8230
},
{
"epoch": 0.5917839701235277,
"grad_norm": 1.0575812171132415,
"learning_rate": 2e-05,
"loss": 0.4992171287536621,
"step": 8240
},
{
"epoch": 0.5925021545532893,
"grad_norm": 0.649065795519634,
"learning_rate": 2e-05,
"loss": 0.48657522201538084,
"step": 8250
},
{
"epoch": 0.5932203389830508,
"grad_norm": 0.7419057419930403,
"learning_rate": 2e-05,
"loss": 0.5374475002288819,
"step": 8260
},
{
"epoch": 0.5939385234128124,
"grad_norm": 0.7161160785946615,
"learning_rate": 2e-05,
"loss": 0.4776909351348877,
"step": 8270
},
{
"epoch": 0.594656707842574,
"grad_norm": 0.7941603240273439,
"learning_rate": 2e-05,
"loss": 0.49909296035766604,
"step": 8280
},
{
"epoch": 0.5953748922723355,
"grad_norm": 0.6930914281999472,
"learning_rate": 2e-05,
"loss": 0.5133593559265137,
"step": 8290
},
{
"epoch": 0.5960930767020971,
"grad_norm": 0.8264045263853194,
"learning_rate": 2e-05,
"loss": 0.5228574275970459,
"step": 8300
},
{
"epoch": 0.5968112611318587,
"grad_norm": 0.7166364806894724,
"learning_rate": 2e-05,
"loss": 0.5094347476959229,
"step": 8310
},
{
"epoch": 0.5975294455616202,
"grad_norm": 0.9883547935157848,
"learning_rate": 2e-05,
"loss": 0.507183313369751,
"step": 8320
},
{
"epoch": 0.5982476299913818,
"grad_norm": 0.6818045790302899,
"learning_rate": 2e-05,
"loss": 0.5156293869018554,
"step": 8330
},
{
"epoch": 0.5989658144211434,
"grad_norm": 1.1649274923193993,
"learning_rate": 2e-05,
"loss": 0.4926125526428223,
"step": 8340
},
{
"epoch": 0.599683998850905,
"grad_norm": 0.8950602167095826,
"learning_rate": 2e-05,
"loss": 0.5391737937927246,
"step": 8350
},
{
"epoch": 0.6004021832806665,
"grad_norm": 0.8032720321985906,
"learning_rate": 2e-05,
"loss": 0.49005904197692873,
"step": 8360
},
{
"epoch": 0.601120367710428,
"grad_norm": 0.9444114287141325,
"learning_rate": 2e-05,
"loss": 0.5007988929748535,
"step": 8370
},
{
"epoch": 0.6018385521401896,
"grad_norm": 1.3404293584256197,
"learning_rate": 2e-05,
"loss": 0.5329407215118408,
"step": 8380
},
{
"epoch": 0.6025567365699511,
"grad_norm": 0.6940813726051713,
"learning_rate": 2e-05,
"loss": 0.491253662109375,
"step": 8390
},
{
"epoch": 0.6032749209997127,
"grad_norm": 1.2727263126023134,
"learning_rate": 2e-05,
"loss": 0.516472339630127,
"step": 8400
},
{
"epoch": 0.6039931054294743,
"grad_norm": 1.0278512869401406,
"learning_rate": 2e-05,
"loss": 0.504970932006836,
"step": 8410
},
{
"epoch": 0.6047112898592358,
"grad_norm": 0.6870247576511328,
"learning_rate": 2e-05,
"loss": 0.48597140312194825,
"step": 8420
},
{
"epoch": 0.6054294742889974,
"grad_norm": 4.959722156118949,
"learning_rate": 2e-05,
"loss": 0.5269383430480957,
"step": 8430
},
{
"epoch": 0.606147658718759,
"grad_norm": 0.7093208420050612,
"learning_rate": 2e-05,
"loss": 0.5159483909606933,
"step": 8440
},
{
"epoch": 0.6068658431485205,
"grad_norm": 0.9311442733522337,
"learning_rate": 2e-05,
"loss": 0.5023585796356201,
"step": 8450
},
{
"epoch": 0.6075840275782821,
"grad_norm": 0.881069301419876,
"learning_rate": 2e-05,
"loss": 0.5063649654388428,
"step": 8460
},
{
"epoch": 0.6083022120080437,
"grad_norm": 1.1942735820623454,
"learning_rate": 2e-05,
"loss": 0.521247386932373,
"step": 8470
},
{
"epoch": 0.6090203964378053,
"grad_norm": 1.2943466084079418,
"learning_rate": 2e-05,
"loss": 0.5363234519958496,
"step": 8480
},
{
"epoch": 0.6097385808675668,
"grad_norm": 0.8888790867392223,
"learning_rate": 2e-05,
"loss": 0.5207231044769287,
"step": 8490
},
{
"epoch": 0.6104567652973284,
"grad_norm": 1.0214070149218373,
"learning_rate": 2e-05,
"loss": 0.513697338104248,
"step": 8500
},
{
"epoch": 0.61117494972709,
"grad_norm": 1.7129271531135648,
"learning_rate": 2e-05,
"loss": 0.5206613540649414,
"step": 8510
},
{
"epoch": 0.6118931341568514,
"grad_norm": 0.819691471954568,
"learning_rate": 2e-05,
"loss": 0.5041644096374511,
"step": 8520
},
{
"epoch": 0.612611318586613,
"grad_norm": 1.0714389709140177,
"learning_rate": 2e-05,
"loss": 0.5190259933471679,
"step": 8530
},
{
"epoch": 0.6133295030163746,
"grad_norm": 0.8630129876200808,
"learning_rate": 2e-05,
"loss": 0.5112534523010254,
"step": 8540
},
{
"epoch": 0.6140476874461361,
"grad_norm": 0.8219783934722581,
"learning_rate": 2e-05,
"loss": 0.540252685546875,
"step": 8550
},
{
"epoch": 0.6147658718758977,
"grad_norm": 0.6470531108432588,
"learning_rate": 2e-05,
"loss": 0.4834305763244629,
"step": 8560
},
{
"epoch": 0.6154840563056593,
"grad_norm": 0.8226758294172642,
"learning_rate": 2e-05,
"loss": 0.4972515106201172,
"step": 8570
},
{
"epoch": 0.6162022407354208,
"grad_norm": 0.8206247692140018,
"learning_rate": 2e-05,
"loss": 0.5131832122802734,
"step": 8580
},
{
"epoch": 0.6169204251651824,
"grad_norm": 62.16203832004988,
"learning_rate": 2e-05,
"loss": 0.5193436622619629,
"step": 8590
},
{
"epoch": 0.617638609594944,
"grad_norm": 1.2695882903308522,
"learning_rate": 2e-05,
"loss": 0.5322408676147461,
"step": 8600
},
{
"epoch": 0.6183567940247056,
"grad_norm": 0.9076095462112073,
"learning_rate": 2e-05,
"loss": 0.5160855770111084,
"step": 8610
},
{
"epoch": 0.6190749784544671,
"grad_norm": 0.8406280564872765,
"learning_rate": 2e-05,
"loss": 0.5171470642089844,
"step": 8620
},
{
"epoch": 0.6197931628842287,
"grad_norm": 2.498870485069704,
"learning_rate": 2e-05,
"loss": 0.512457799911499,
"step": 8630
},
{
"epoch": 0.6205113473139903,
"grad_norm": 1.226644039627787,
"learning_rate": 2e-05,
"loss": 0.5022732734680175,
"step": 8640
},
{
"epoch": 0.6212295317437518,
"grad_norm": 0.8657902462792295,
"learning_rate": 2e-05,
"loss": 0.49957880973815916,
"step": 8650
},
{
"epoch": 0.6219477161735134,
"grad_norm": 0.7679039644878018,
"learning_rate": 2e-05,
"loss": 0.5004348278045654,
"step": 8660
},
{
"epoch": 0.6226659006032749,
"grad_norm": 0.7782442962805688,
"learning_rate": 2e-05,
"loss": 0.5256059169769287,
"step": 8670
},
{
"epoch": 0.6233840850330364,
"grad_norm": 0.9964554303751821,
"learning_rate": 2e-05,
"loss": 0.5036609649658204,
"step": 8680
},
{
"epoch": 0.624102269462798,
"grad_norm": 0.8596457806783621,
"learning_rate": 2e-05,
"loss": 0.4924436569213867,
"step": 8690
},
{
"epoch": 0.6248204538925596,
"grad_norm": 0.6805149003310149,
"learning_rate": 2e-05,
"loss": 0.5287024497985839,
"step": 8700
},
{
"epoch": 0.6255386383223211,
"grad_norm": 0.9562927276609984,
"learning_rate": 2e-05,
"loss": 0.520150089263916,
"step": 8710
},
{
"epoch": 0.6262568227520827,
"grad_norm": 1.3048826357813843,
"learning_rate": 2e-05,
"loss": 0.5008325099945068,
"step": 8720
},
{
"epoch": 0.6269750071818443,
"grad_norm": 1.1084604731444463,
"learning_rate": 2e-05,
"loss": 0.5092177391052246,
"step": 8730
},
{
"epoch": 0.6276931916116059,
"grad_norm": 10.060204228180487,
"learning_rate": 2e-05,
"loss": 0.5027655601501465,
"step": 8740
},
{
"epoch": 0.6284113760413674,
"grad_norm": 0.7200853359435939,
"learning_rate": 2e-05,
"loss": 0.5059186935424804,
"step": 8750
},
{
"epoch": 0.629129560471129,
"grad_norm": 0.832765197337854,
"learning_rate": 2e-05,
"loss": 0.5056815147399902,
"step": 8760
},
{
"epoch": 0.6298477449008906,
"grad_norm": 0.7755219762976564,
"learning_rate": 2e-05,
"loss": 0.5105802059173584,
"step": 8770
},
{
"epoch": 0.6305659293306521,
"grad_norm": 1.002964651112839,
"learning_rate": 2e-05,
"loss": 0.5142621994018555,
"step": 8780
},
{
"epoch": 0.6312841137604137,
"grad_norm": 0.8674991260143492,
"learning_rate": 2e-05,
"loss": 0.5357322216033935,
"step": 8790
},
{
"epoch": 0.6320022981901753,
"grad_norm": 198.96674989277645,
"learning_rate": 2e-05,
"loss": 0.5395197868347168,
"step": 8800
},
{
"epoch": 0.6327204826199369,
"grad_norm": 2.062227684505636,
"learning_rate": 2e-05,
"loss": 0.537823486328125,
"step": 8810
},
{
"epoch": 0.6334386670496983,
"grad_norm": 1.1324150403991549,
"learning_rate": 2e-05,
"loss": 0.5076553344726562,
"step": 8820
},
{
"epoch": 0.6341568514794599,
"grad_norm": 1.0870709003065333,
"learning_rate": 2e-05,
"loss": 0.5182052612304687,
"step": 8830
},
{
"epoch": 0.6348750359092215,
"grad_norm": 1.0791431406496188,
"learning_rate": 2e-05,
"loss": 0.5153659820556641,
"step": 8840
},
{
"epoch": 0.635593220338983,
"grad_norm": 0.8558117762286055,
"learning_rate": 2e-05,
"loss": 0.4925567626953125,
"step": 8850
},
{
"epoch": 0.6363114047687446,
"grad_norm": 1.1882192262264077,
"learning_rate": 2e-05,
"loss": 0.5101389408111572,
"step": 8860
},
{
"epoch": 0.6370295891985062,
"grad_norm": 6.63035602862505,
"learning_rate": 2e-05,
"loss": 0.5008525848388672,
"step": 8870
},
{
"epoch": 0.6377477736282677,
"grad_norm": 0.9050831776485863,
"learning_rate": 2e-05,
"loss": 0.5199275970458984,
"step": 8880
},
{
"epoch": 0.6384659580580293,
"grad_norm": 0.964999852102051,
"learning_rate": 2e-05,
"loss": 0.5129219055175781,
"step": 8890
},
{
"epoch": 0.6391841424877909,
"grad_norm": 2.3721042704921036,
"learning_rate": 2e-05,
"loss": 0.534566593170166,
"step": 8900
},
{
"epoch": 0.6399023269175524,
"grad_norm": 0.7700269134295917,
"learning_rate": 2e-05,
"loss": 0.46931166648864747,
"step": 8910
},
{
"epoch": 0.640620511347314,
"grad_norm": 8.167039330136914,
"learning_rate": 2e-05,
"loss": 0.488972806930542,
"step": 8920
},
{
"epoch": 0.6413386957770756,
"grad_norm": 0.7535600833819553,
"learning_rate": 2e-05,
"loss": 0.5017453670501709,
"step": 8930
},
{
"epoch": 0.6420568802068372,
"grad_norm": 0.9904951154645321,
"learning_rate": 2e-05,
"loss": 0.5064147472381592,
"step": 8940
},
{
"epoch": 0.6427750646365987,
"grad_norm": 0.7859821892119616,
"learning_rate": 2e-05,
"loss": 0.520528507232666,
"step": 8950
},
{
"epoch": 0.6434932490663603,
"grad_norm": 1.0431235432974566,
"learning_rate": 2e-05,
"loss": 0.49713592529296874,
"step": 8960
},
{
"epoch": 0.6442114334961218,
"grad_norm": 0.8395028767068718,
"learning_rate": 2e-05,
"loss": 0.5224717140197754,
"step": 8970
},
{
"epoch": 0.6449296179258833,
"grad_norm": 0.694188100245605,
"learning_rate": 2e-05,
"loss": 0.5168360710144043,
"step": 8980
},
{
"epoch": 0.6456478023556449,
"grad_norm": 1.068225716861055,
"learning_rate": 2e-05,
"loss": 0.5143491744995117,
"step": 8990
},
{
"epoch": 0.6463659867854065,
"grad_norm": 0.8541029449859785,
"learning_rate": 2e-05,
"loss": 0.5103260040283203,
"step": 9000
},
{
"epoch": 0.647084171215168,
"grad_norm": 0.8690543095361802,
"learning_rate": 2e-05,
"loss": 0.5237655162811279,
"step": 9010
},
{
"epoch": 0.6478023556449296,
"grad_norm": 0.8467285761476944,
"learning_rate": 2e-05,
"loss": 0.5001975059509277,
"step": 9020
},
{
"epoch": 0.6485205400746912,
"grad_norm": 7.525563015474066,
"learning_rate": 2e-05,
"loss": 0.4840196132659912,
"step": 9030
},
{
"epoch": 0.6492387245044527,
"grad_norm": 0.6835468654730276,
"learning_rate": 2e-05,
"loss": 0.5143831253051758,
"step": 9040
},
{
"epoch": 0.6499569089342143,
"grad_norm": 0.8037601220370326,
"learning_rate": 2e-05,
"loss": 0.5020384311676025,
"step": 9050
},
{
"epoch": 0.6506750933639759,
"grad_norm": 0.9312423195333381,
"learning_rate": 2e-05,
"loss": 0.534348440170288,
"step": 9060
},
{
"epoch": 0.6513932777937375,
"grad_norm": 0.9173742808652284,
"learning_rate": 2e-05,
"loss": 0.49380884170532224,
"step": 9070
},
{
"epoch": 0.652111462223499,
"grad_norm": 0.7855778967376087,
"learning_rate": 2e-05,
"loss": 0.4893780708312988,
"step": 9080
},
{
"epoch": 0.6528296466532606,
"grad_norm": 0.7215133394347666,
"learning_rate": 2e-05,
"loss": 0.4967243194580078,
"step": 9090
},
{
"epoch": 0.6535478310830222,
"grad_norm": 0.8159180236118575,
"learning_rate": 2e-05,
"loss": 0.48772449493408204,
"step": 9100
},
{
"epoch": 0.6542660155127836,
"grad_norm": 0.9550913030403764,
"learning_rate": 2e-05,
"loss": 0.4825115203857422,
"step": 9110
},
{
"epoch": 0.6549841999425452,
"grad_norm": 0.7343105764502756,
"learning_rate": 2e-05,
"loss": 0.5131556510925293,
"step": 9120
},
{
"epoch": 0.6557023843723068,
"grad_norm": 0.6908879276202479,
"learning_rate": 2e-05,
"loss": 0.4985659122467041,
"step": 9130
},
{
"epoch": 0.6564205688020683,
"grad_norm": 1.0046669605780867,
"learning_rate": 2e-05,
"loss": 0.5008234977722168,
"step": 9140
},
{
"epoch": 0.6571387532318299,
"grad_norm": 0.8009316989748421,
"learning_rate": 2e-05,
"loss": 0.4895171642303467,
"step": 9150
},
{
"epoch": 0.6578569376615915,
"grad_norm": 0.8574310178694184,
"learning_rate": 2e-05,
"loss": 0.5046119689941406,
"step": 9160
},
{
"epoch": 0.658575122091353,
"grad_norm": 0.9421000902216401,
"learning_rate": 2e-05,
"loss": 0.5011324405670166,
"step": 9170
},
{
"epoch": 0.6592933065211146,
"grad_norm": 0.8227846951714585,
"learning_rate": 2e-05,
"loss": 0.49074587821960447,
"step": 9180
},
{
"epoch": 0.6600114909508762,
"grad_norm": 1.2268734002614736,
"learning_rate": 2e-05,
"loss": 0.5051110267639161,
"step": 9190
},
{
"epoch": 0.6607296753806378,
"grad_norm": 1.0741153953172702,
"learning_rate": 2e-05,
"loss": 0.5212117671966553,
"step": 9200
},
{
"epoch": 0.6614478598103993,
"grad_norm": 0.8336769392844531,
"learning_rate": 2e-05,
"loss": 0.504359245300293,
"step": 9210
},
{
"epoch": 0.6621660442401609,
"grad_norm": 0.8039499594644233,
"learning_rate": 2e-05,
"loss": 0.5139644145965576,
"step": 9220
},
{
"epoch": 0.6628842286699225,
"grad_norm": 1.3479138978430436,
"learning_rate": 2e-05,
"loss": 0.4929664134979248,
"step": 9230
},
{
"epoch": 0.663602413099684,
"grad_norm": 4.178080049579891,
"learning_rate": 2e-05,
"loss": 0.48785648345947263,
"step": 9240
},
{
"epoch": 0.6643205975294456,
"grad_norm": 0.7741917884392573,
"learning_rate": 2e-05,
"loss": 0.4977989196777344,
"step": 9250
},
{
"epoch": 0.6650387819592071,
"grad_norm": 0.7876722591355765,
"learning_rate": 2e-05,
"loss": 0.5124330043792724,
"step": 9260
},
{
"epoch": 0.6657569663889686,
"grad_norm": 0.7911563466199747,
"learning_rate": 2e-05,
"loss": 0.5316314220428466,
"step": 9270
},
{
"epoch": 0.6664751508187302,
"grad_norm": 0.7646048377256145,
"learning_rate": 2e-05,
"loss": 0.4930027961730957,
"step": 9280
},
{
"epoch": 0.6671933352484918,
"grad_norm": 1.1538268855058178,
"learning_rate": 2e-05,
"loss": 0.5132038116455078,
"step": 9290
},
{
"epoch": 0.6679115196782534,
"grad_norm": 1.31892730450197,
"learning_rate": 2e-05,
"loss": 0.5060904502868653,
"step": 9300
},
{
"epoch": 0.6686297041080149,
"grad_norm": 0.8125488573743432,
"learning_rate": 2e-05,
"loss": 0.49868125915527345,
"step": 9310
},
{
"epoch": 0.6693478885377765,
"grad_norm": 0.7003192147336307,
"learning_rate": 2e-05,
"loss": 0.5043570518493652,
"step": 9320
},
{
"epoch": 0.6700660729675381,
"grad_norm": 0.7542089601901899,
"learning_rate": 2e-05,
"loss": 0.5118447780609131,
"step": 9330
},
{
"epoch": 0.6707842573972996,
"grad_norm": 0.7220518428832727,
"learning_rate": 2e-05,
"loss": 0.5081843376159668,
"step": 9340
},
{
"epoch": 0.6715024418270612,
"grad_norm": 0.7132719008194287,
"learning_rate": 2e-05,
"loss": 0.5141977310180664,
"step": 9350
},
{
"epoch": 0.6722206262568228,
"grad_norm": 0.8187081831329146,
"learning_rate": 2e-05,
"loss": 0.5251072883605957,
"step": 9360
},
{
"epoch": 0.6729388106865843,
"grad_norm": 0.7422764646525389,
"learning_rate": 2e-05,
"loss": 0.5174417972564698,
"step": 9370
},
{
"epoch": 0.6736569951163459,
"grad_norm": 0.6772090663468322,
"learning_rate": 2e-05,
"loss": 0.5210283756256103,
"step": 9380
},
{
"epoch": 0.6743751795461075,
"grad_norm": 0.9541678286018351,
"learning_rate": 2e-05,
"loss": 0.48545279502868655,
"step": 9390
},
{
"epoch": 0.675093363975869,
"grad_norm": 0.6473299302647226,
"learning_rate": 2e-05,
"loss": 0.4997104167938232,
"step": 9400
},
{
"epoch": 0.6758115484056305,
"grad_norm": 0.8292799263387656,
"learning_rate": 2e-05,
"loss": 0.5372878074645996,
"step": 9410
},
{
"epoch": 0.6765297328353921,
"grad_norm": 0.7721712136673651,
"learning_rate": 2e-05,
"loss": 0.4841670036315918,
"step": 9420
},
{
"epoch": 0.6772479172651537,
"grad_norm": 0.6648753130399344,
"learning_rate": 2e-05,
"loss": 0.5299575805664063,
"step": 9430
},
{
"epoch": 0.6779661016949152,
"grad_norm": 0.8437289078671946,
"learning_rate": 2e-05,
"loss": 0.4855644226074219,
"step": 9440
},
{
"epoch": 0.6786842861246768,
"grad_norm": 0.6969892609214623,
"learning_rate": 2e-05,
"loss": 0.49136805534362793,
"step": 9450
},
{
"epoch": 0.6794024705544384,
"grad_norm": 0.8113969194869025,
"learning_rate": 2e-05,
"loss": 0.5266856193542481,
"step": 9460
},
{
"epoch": 0.6801206549841999,
"grad_norm": 0.7162702207492141,
"learning_rate": 2e-05,
"loss": 0.48981504440307616,
"step": 9470
},
{
"epoch": 0.6808388394139615,
"grad_norm": 0.6210401677366795,
"learning_rate": 2e-05,
"loss": 0.4900937557220459,
"step": 9480
},
{
"epoch": 0.6815570238437231,
"grad_norm": 0.7457630363043171,
"learning_rate": 2e-05,
"loss": 0.5247382164001465,
"step": 9490
},
{
"epoch": 0.6822752082734846,
"grad_norm": 0.9108420203250669,
"learning_rate": 2e-05,
"loss": 0.494512939453125,
"step": 9500
},
{
"epoch": 0.6829933927032462,
"grad_norm": 1.2558303017041743,
"learning_rate": 2e-05,
"loss": 0.49795827865600584,
"step": 9510
},
{
"epoch": 0.6837115771330078,
"grad_norm": 0.8507671724869936,
"learning_rate": 2e-05,
"loss": 0.5365367889404297,
"step": 9520
},
{
"epoch": 0.6844297615627694,
"grad_norm": 1.0073260061762628,
"learning_rate": 2e-05,
"loss": 0.49969139099121096,
"step": 9530
},
{
"epoch": 0.6851479459925309,
"grad_norm": 0.7603100911694591,
"learning_rate": 2e-05,
"loss": 0.5182240962982178,
"step": 9540
},
{
"epoch": 0.6858661304222925,
"grad_norm": 0.9424341276132211,
"learning_rate": 2e-05,
"loss": 0.544720983505249,
"step": 9550
},
{
"epoch": 0.686584314852054,
"grad_norm": 0.6737841884766598,
"learning_rate": 2e-05,
"loss": 0.5088635921478272,
"step": 9560
},
{
"epoch": 0.6873024992818155,
"grad_norm": 0.7549138707067549,
"learning_rate": 2e-05,
"loss": 0.4787275314331055,
"step": 9570
},
{
"epoch": 0.6880206837115771,
"grad_norm": 0.7101385192214543,
"learning_rate": 2e-05,
"loss": 0.5010103225708008,
"step": 9580
},
{
"epoch": 0.6887388681413387,
"grad_norm": 1.1474292443642284,
"learning_rate": 2e-05,
"loss": 0.4837817192077637,
"step": 9590
},
{
"epoch": 0.6894570525711002,
"grad_norm": 0.7013246529164334,
"learning_rate": 2e-05,
"loss": 0.4819814205169678,
"step": 9600
},
{
"epoch": 0.6901752370008618,
"grad_norm": 0.6529390742871395,
"learning_rate": 2e-05,
"loss": 0.5086185932159424,
"step": 9610
},
{
"epoch": 0.6908934214306234,
"grad_norm": 0.7782710136618118,
"learning_rate": 2e-05,
"loss": 0.4872158050537109,
"step": 9620
},
{
"epoch": 0.691611605860385,
"grad_norm": 1.1222326342809337,
"learning_rate": 2e-05,
"loss": 0.491532564163208,
"step": 9630
},
{
"epoch": 0.6923297902901465,
"grad_norm": 0.7022236402471518,
"learning_rate": 2e-05,
"loss": 0.49276485443115237,
"step": 9640
},
{
"epoch": 0.6930479747199081,
"grad_norm": 0.7671296185623273,
"learning_rate": 2e-05,
"loss": 0.5147552967071534,
"step": 9650
},
{
"epoch": 0.6937661591496697,
"grad_norm": 0.6736593638174492,
"learning_rate": 2e-05,
"loss": 0.4832132816314697,
"step": 9660
},
{
"epoch": 0.6944843435794312,
"grad_norm": 0.7417349674796883,
"learning_rate": 2e-05,
"loss": 0.5162530899047851,
"step": 9670
},
{
"epoch": 0.6952025280091928,
"grad_norm": 0.8602369040568648,
"learning_rate": 2e-05,
"loss": 0.4703378677368164,
"step": 9680
},
{
"epoch": 0.6959207124389544,
"grad_norm": 0.6729828152110714,
"learning_rate": 2e-05,
"loss": 0.5032936573028565,
"step": 9690
},
{
"epoch": 0.6966388968687159,
"grad_norm": 0.7677134240726674,
"learning_rate": 2e-05,
"loss": 0.5120569229125976,
"step": 9700
},
{
"epoch": 0.6973570812984774,
"grad_norm": 0.6510157646248721,
"learning_rate": 2e-05,
"loss": 0.5005989074707031,
"step": 9710
},
{
"epoch": 0.698075265728239,
"grad_norm": 0.8041225085750044,
"learning_rate": 2e-05,
"loss": 0.4983852386474609,
"step": 9720
},
{
"epoch": 0.6987934501580005,
"grad_norm": 0.650459642479386,
"learning_rate": 2e-05,
"loss": 0.49165830612182615,
"step": 9730
},
{
"epoch": 0.6995116345877621,
"grad_norm": 0.9283370805219022,
"learning_rate": 2e-05,
"loss": 0.513530969619751,
"step": 9740
},
{
"epoch": 0.7002298190175237,
"grad_norm": 0.7157969345529395,
"learning_rate": 2e-05,
"loss": 0.5168548107147217,
"step": 9750
},
{
"epoch": 0.7009480034472853,
"grad_norm": 0.6974719754083568,
"learning_rate": 2e-05,
"loss": 0.48286094665527346,
"step": 9760
},
{
"epoch": 0.7016661878770468,
"grad_norm": 0.8087386876515639,
"learning_rate": 2e-05,
"loss": 0.539850902557373,
"step": 9770
},
{
"epoch": 0.7023843723068084,
"grad_norm": 0.8449486926334513,
"learning_rate": 2e-05,
"loss": 0.5181422233581543,
"step": 9780
},
{
"epoch": 0.70310255673657,
"grad_norm": 0.8246841499288998,
"learning_rate": 2e-05,
"loss": 0.5296232223510742,
"step": 9790
},
{
"epoch": 0.7038207411663315,
"grad_norm": 0.8441972113279701,
"learning_rate": 2e-05,
"loss": 0.5138182640075684,
"step": 9800
},
{
"epoch": 0.7045389255960931,
"grad_norm": 0.6978220936624083,
"learning_rate": 2e-05,
"loss": 0.504146957397461,
"step": 9810
},
{
"epoch": 0.7052571100258547,
"grad_norm": 0.9538271965673942,
"learning_rate": 2e-05,
"loss": 0.4930141448974609,
"step": 9820
},
{
"epoch": 0.7059752944556162,
"grad_norm": 0.9674184671773415,
"learning_rate": 2e-05,
"loss": 0.5063184261322021,
"step": 9830
},
{
"epoch": 0.7066934788853778,
"grad_norm": 0.6763419509796172,
"learning_rate": 2e-05,
"loss": 0.4898832321166992,
"step": 9840
},
{
"epoch": 0.7074116633151394,
"grad_norm": 0.956905520975154,
"learning_rate": 2e-05,
"loss": 0.5021014213562012,
"step": 9850
},
{
"epoch": 0.7081298477449008,
"grad_norm": 2.270447912664535,
"learning_rate": 2e-05,
"loss": 0.4822347164154053,
"step": 9860
},
{
"epoch": 0.7088480321746624,
"grad_norm": 1.217035834652496,
"learning_rate": 2e-05,
"loss": 0.4750885009765625,
"step": 9870
},
{
"epoch": 0.709566216604424,
"grad_norm": 0.743298974018276,
"learning_rate": 2e-05,
"loss": 0.48802480697631834,
"step": 9880
},
{
"epoch": 0.7102844010341856,
"grad_norm": 0.8609715871464494,
"learning_rate": 2e-05,
"loss": 0.5022787094116211,
"step": 9890
},
{
"epoch": 0.7110025854639471,
"grad_norm": 0.7718203351017588,
"learning_rate": 2e-05,
"loss": 0.4943395614624023,
"step": 9900
},
{
"epoch": 0.7117207698937087,
"grad_norm": 0.8118412687476312,
"learning_rate": 2e-05,
"loss": 0.4928037166595459,
"step": 9910
},
{
"epoch": 0.7124389543234703,
"grad_norm": 0.9673593109876247,
"learning_rate": 2e-05,
"loss": 0.5028488636016846,
"step": 9920
},
{
"epoch": 0.7131571387532318,
"grad_norm": 0.7358867500986613,
"learning_rate": 2e-05,
"loss": 0.4998504638671875,
"step": 9930
},
{
"epoch": 0.7138753231829934,
"grad_norm": 0.7010175902285685,
"learning_rate": 2e-05,
"loss": 0.5135861396789551,
"step": 9940
},
{
"epoch": 0.714593507612755,
"grad_norm": 0.6837855684246692,
"learning_rate": 2e-05,
"loss": 0.47791147232055664,
"step": 9950
},
{
"epoch": 0.7153116920425165,
"grad_norm": 0.741899840176516,
"learning_rate": 2e-05,
"loss": 0.48313093185424805,
"step": 9960
},
{
"epoch": 0.7160298764722781,
"grad_norm": 0.7418795054544963,
"learning_rate": 2e-05,
"loss": 0.4975980281829834,
"step": 9970
},
{
"epoch": 0.7167480609020397,
"grad_norm": 0.7137047895667901,
"learning_rate": 2e-05,
"loss": 0.48713226318359376,
"step": 9980
},
{
"epoch": 0.7174662453318013,
"grad_norm": 0.6453566510833598,
"learning_rate": 2e-05,
"loss": 0.4967761516571045,
"step": 9990
},
{
"epoch": 0.7181844297615627,
"grad_norm": 0.7589838223873897,
"learning_rate": 2e-05,
"loss": 0.4966542720794678,
"step": 10000
},
{
"epoch": 0.7189026141913243,
"grad_norm": 1.561311475925959,
"learning_rate": 2e-05,
"loss": 0.4773681640625,
"step": 10010
},
{
"epoch": 0.7196207986210859,
"grad_norm": 0.9214936780931193,
"learning_rate": 2e-05,
"loss": 0.49508371353149416,
"step": 10020
},
{
"epoch": 0.7203389830508474,
"grad_norm": 0.7295388779420731,
"learning_rate": 2e-05,
"loss": 0.47771902084350587,
"step": 10030
},
{
"epoch": 0.721057167480609,
"grad_norm": 1.0743039233004235,
"learning_rate": 2e-05,
"loss": 0.5382555961608887,
"step": 10040
},
{
"epoch": 0.7217753519103706,
"grad_norm": 0.8305238133633193,
"learning_rate": 2e-05,
"loss": 0.5206102848052978,
"step": 10050
},
{
"epoch": 0.7224935363401321,
"grad_norm": 0.8811007354794976,
"learning_rate": 2e-05,
"loss": 0.5061130046844482,
"step": 10060
},
{
"epoch": 0.7232117207698937,
"grad_norm": 0.730241845819166,
"learning_rate": 2e-05,
"loss": 0.49259276390075685,
"step": 10070
},
{
"epoch": 0.7239299051996553,
"grad_norm": 1.3890467532318163,
"learning_rate": 2e-05,
"loss": 0.4953322887420654,
"step": 10080
},
{
"epoch": 0.7246480896294168,
"grad_norm": 0.8605629665675073,
"learning_rate": 2e-05,
"loss": 0.497589635848999,
"step": 10090
},
{
"epoch": 0.7253662740591784,
"grad_norm": 0.7707076683685329,
"learning_rate": 2e-05,
"loss": 0.4850158214569092,
"step": 10100
},
{
"epoch": 0.72608445848894,
"grad_norm": 0.8057737709605227,
"learning_rate": 2e-05,
"loss": 0.5092198371887207,
"step": 10110
},
{
"epoch": 0.7268026429187016,
"grad_norm": 0.7602429773419251,
"learning_rate": 2e-05,
"loss": 0.4953362464904785,
"step": 10120
},
{
"epoch": 0.7275208273484631,
"grad_norm": 0.9004351263887387,
"learning_rate": 2e-05,
"loss": 0.5082661151885987,
"step": 10130
},
{
"epoch": 0.7282390117782247,
"grad_norm": 1.023533183060986,
"learning_rate": 2e-05,
"loss": 0.5238260269165039,
"step": 10140
},
{
"epoch": 0.7289571962079862,
"grad_norm": 0.6108981484038831,
"learning_rate": 2e-05,
"loss": 0.4772487640380859,
"step": 10150
},
{
"epoch": 0.7296753806377477,
"grad_norm": 0.8470093941472545,
"learning_rate": 2e-05,
"loss": 0.4819645881652832,
"step": 10160
},
{
"epoch": 0.7303935650675093,
"grad_norm": 0.7038705942450684,
"learning_rate": 2e-05,
"loss": 0.5145659923553467,
"step": 10170
},
{
"epoch": 0.7311117494972709,
"grad_norm": 1.006397114035138,
"learning_rate": 1.999957274187931e-05,
"loss": 0.49317407608032227,
"step": 10180
},
{
"epoch": 0.7318299339270324,
"grad_norm": 0.7225558117387574,
"learning_rate": 1.9991978062682906e-05,
"loss": 0.5225354194641113,
"step": 10190
},
{
"epoch": 0.732548118356794,
"grad_norm": 1.1971314116748577,
"learning_rate": 1.997489709987378e-05,
"loss": 0.5093063354492188,
"step": 10200
},
{
"epoch": 0.7332663027865556,
"grad_norm": 0.7779179461104959,
"learning_rate": 1.9948346151502135e-05,
"loss": 0.5011675357818604,
"step": 10210
},
{
"epoch": 0.7339844872163171,
"grad_norm": 0.8959382686869318,
"learning_rate": 1.991235055154325e-05,
"loss": 0.4797985076904297,
"step": 10220
},
{
"epoch": 0.7347026716460787,
"grad_norm": 0.6683902386248717,
"learning_rate": 1.9866944645724693e-05,
"loss": 0.488443660736084,
"step": 10230
},
{
"epoch": 0.7354208560758403,
"grad_norm": 0.8615015451444095,
"learning_rate": 1.9812171758754842e-05,
"loss": 0.49929037094116213,
"step": 10240
},
{
"epoch": 0.7361390405056019,
"grad_norm": 1.0959727737609588,
"learning_rate": 1.9748084152983987e-05,
"loss": 0.5262956619262695,
"step": 10250
},
{
"epoch": 0.7368572249353634,
"grad_norm": 0.7986082458875614,
"learning_rate": 1.967474297853744e-05,
"loss": 0.5348004341125489,
"step": 10260
},
{
"epoch": 0.737575409365125,
"grad_norm": 0.770359901377329,
"learning_rate": 1.9592218214968256e-05,
"loss": 0.4844058036804199,
"step": 10270
},
{
"epoch": 0.7382935937948866,
"grad_norm": 0.7161493893989926,
"learning_rate": 1.9500588604485213e-05,
"loss": 0.4777571678161621,
"step": 10280
},
{
"epoch": 0.7390117782246481,
"grad_norm": 0.7840463267378235,
"learning_rate": 1.939994157681979e-05,
"loss": 0.5069754600524903,
"step": 10290
},
{
"epoch": 0.7397299626544096,
"grad_norm": 0.8702085806472359,
"learning_rate": 1.9290373165803803e-05,
"loss": 0.5046972751617431,
"step": 10300
},
{
"epoch": 0.7404481470841712,
"grad_norm": 0.7275044406189317,
"learning_rate": 1.9171987917737328e-05,
"loss": 0.49483866691589357,
"step": 10310
},
{
"epoch": 0.7411663315139327,
"grad_norm": 0.834537313964422,
"learning_rate": 1.9044898791634322e-05,
"loss": 0.5052151203155517,
"step": 10320
},
{
"epoch": 0.7418845159436943,
"grad_norm": 0.7228585176787719,
"learning_rate": 1.8909227051441126e-05,
"loss": 0.5003730773925781,
"step": 10330
},
{
"epoch": 0.7426027003734559,
"grad_norm": 0.6379062259088754,
"learning_rate": 1.8765102150330692e-05,
"loss": 0.48312959671020506,
"step": 10340
},
{
"epoch": 0.7433208848032175,
"grad_norm": 0.764896927942744,
"learning_rate": 1.861266160718297e-05,
"loss": 0.5113920211791992,
"step": 10350
},
{
"epoch": 0.744039069232979,
"grad_norm": 0.7177160333846159,
"learning_rate": 1.845205087536923e-05,
"loss": 0.49533801078796386,
"step": 10360
},
{
"epoch": 0.7447572536627406,
"grad_norm": 0.8876923639434253,
"learning_rate": 1.8283423203965633e-05,
"loss": 0.480242919921875,
"step": 10370
},
{
"epoch": 0.7454754380925022,
"grad_norm": 0.7683825170137376,
"learning_rate": 1.810693949152835e-05,
"loss": 0.46788616180419923,
"step": 10380
},
{
"epoch": 0.7461936225222637,
"grad_norm": 0.6801118859931555,
"learning_rate": 1.7922768132569902e-05,
"loss": 0.48964605331420896,
"step": 10390
},
{
"epoch": 0.7469118069520253,
"grad_norm": 0.6855987272426028,
"learning_rate": 1.773108485688302e-05,
"loss": 0.5022527694702148,
"step": 10400
},
{
"epoch": 0.7476299913817869,
"grad_norm": 0.7524677273366701,
"learning_rate": 1.7532072561865537e-05,
"loss": 0.4923431396484375,
"step": 10410
},
{
"epoch": 0.7483481758115484,
"grad_norm": 0.6399228342497926,
"learning_rate": 1.732592113800616e-05,
"loss": 0.48642377853393554,
"step": 10420
},
{
"epoch": 0.74906636024131,
"grad_norm": 0.6539921439941283,
"learning_rate": 1.7112827287697668e-05,
"loss": 0.47266111373901365,
"step": 10430
},
{
"epoch": 0.7497845446710716,
"grad_norm": 0.8101562024066615,
"learning_rate": 1.6892994337550452e-05,
"loss": 0.5004039287567139,
"step": 10440
},
{
"epoch": 0.750502729100833,
"grad_norm": 0.7738856679991685,
"learning_rate": 1.6666632044385487e-05,
"loss": 0.5102338790893555,
"step": 10450
},
{
"epoch": 0.7512209135305946,
"grad_norm": 0.7097038591552551,
"learning_rate": 1.6433956395091757e-05,
"loss": 0.4951935291290283,
"step": 10460
},
{
"epoch": 0.7519390979603562,
"grad_norm": 1.480443328311553,
"learning_rate": 1.6195189400539238e-05,
"loss": 0.5183979034423828,
"step": 10470
},
{
"epoch": 0.7526572823901178,
"grad_norm": 1.4329786931925914,
"learning_rate": 1.595055888374398e-05,
"loss": 0.4820581912994385,
"step": 10480
},
{
"epoch": 0.7533754668198793,
"grad_norm": 0.8067309501770414,
"learning_rate": 1.5700298262487434e-05,
"loss": 0.4763821601867676,
"step": 10490
},
{
"epoch": 0.7540936512496409,
"grad_norm": 1.6940494587042438,
"learning_rate": 1.5444646326597467e-05,
"loss": 0.4906010150909424,
"step": 10500
},
{
"epoch": 0.7548118356794025,
"grad_norm": 1.4124258679674697,
"learning_rate": 1.5183847010103586e-05,
"loss": 0.5322961807250977,
"step": 10510
},
{
"epoch": 0.755530020109164,
"grad_norm": 0.7309718960979228,
"learning_rate": 1.4918149158483686e-05,
"loss": 0.5117415904998779,
"step": 10520
},
{
"epoch": 0.7562482045389256,
"grad_norm": 0.8962143924103495,
"learning_rate": 1.464780629122452e-05,
"loss": 0.48221621513366697,
"step": 10530
},
{
"epoch": 0.7569663889686872,
"grad_norm": 0.8418727341784377,
"learning_rate": 1.4373076359922374e-05,
"loss": 0.5069468021392822,
"step": 10540
},
{
"epoch": 0.7576845733984487,
"grad_norm": 0.8341503562073556,
"learning_rate": 1.4094221502154774e-05,
"loss": 0.5205643653869629,
"step": 10550
},
{
"epoch": 0.7584027578282103,
"grad_norm": 0.6981225715115282,
"learning_rate": 1.3811507791358113e-05,
"loss": 0.519122314453125,
"step": 10560
},
{
"epoch": 0.7591209422579719,
"grad_norm": 3.3211703039249167,
"learning_rate": 1.3525204982949753e-05,
"loss": 0.47844367027282714,
"step": 10570
},
{
"epoch": 0.7598391266877335,
"grad_norm": 0.9693860064977068,
"learning_rate": 1.3235586256936996e-05,
"loss": 0.511298656463623,
"step": 10580
},
{
"epoch": 0.760557311117495,
"grad_norm": 0.6805860488179284,
"learning_rate": 1.294292795725836e-05,
"loss": 0.46606926918029784,
"step": 10590
},
{
"epoch": 0.7612754955472565,
"grad_norm": 0.6493462221544662,
"learning_rate": 1.2647509328105978e-05,
"loss": 0.46816301345825195,
"step": 10600
},
{
"epoch": 0.7619936799770181,
"grad_norm": 0.8959785926908217,
"learning_rate": 1.2349612247480686e-05,
"loss": 0.4695296287536621,
"step": 10610
},
{
"epoch": 0.7627118644067796,
"grad_norm": 0.7557160837668521,
"learning_rate": 1.2049520958234e-05,
"loss": 0.485294246673584,
"step": 10620
},
{
"epoch": 0.7634300488365412,
"grad_norm": 0.8741772117243284,
"learning_rate": 1.1747521796853638e-05,
"loss": 0.47388572692871095,
"step": 10630
},
{
"epoch": 0.7641482332663028,
"grad_norm": 0.6205168127577962,
"learning_rate": 1.1443902920251395e-05,
"loss": 0.5080687522888183,
"step": 10640
},
{
"epoch": 0.7648664176960643,
"grad_norm": 0.8814148770336598,
"learning_rate": 1.1138954030814009e-05,
"loss": 0.4809291839599609,
"step": 10650
},
{
"epoch": 0.7655846021258259,
"grad_norm": 0.7927920678169721,
"learning_rate": 1.0832966099979388e-05,
"loss": 0.4816349983215332,
"step": 10660
},
{
"epoch": 0.7663027865555875,
"grad_norm": 0.7083922715795367,
"learning_rate": 1.0526231090601969e-05,
"loss": 0.5214847564697266,
"step": 10670
},
{
"epoch": 0.767020970985349,
"grad_norm": 0.6547826686098243,
"learning_rate": 1.0219041678372106e-05,
"loss": 0.4824196815490723,
"step": 10680
},
{
"epoch": 0.7677391554151106,
"grad_norm": 0.8411538722119711,
"learning_rate": 9.911690972555268e-06,
"loss": 0.48386659622192385,
"step": 10690
},
{
"epoch": 0.7684573398448722,
"grad_norm": 1.3832305027352114,
"learning_rate": 9.604472236317541e-06,
"loss": 0.4550494194030762,
"step": 10700
},
{
"epoch": 0.7691755242746338,
"grad_norm": 0.8665870478271425,
"learning_rate": 9.297678606904335e-06,
"loss": 0.4687366008758545,
"step": 10710
},
{
"epoch": 0.7698937087043953,
"grad_norm": 0.9077269033444492,
"learning_rate": 8.991602815939155e-06,
"loss": 0.4746860504150391,
"step": 10720
},
{
"epoch": 0.7706118931341569,
"grad_norm": 0.7666516464407973,
"learning_rate": 8.686536910109464e-06,
"loss": 0.4732991695404053,
"step": 10730
},
{
"epoch": 0.7713300775639184,
"grad_norm": 0.6621814017680565,
"learning_rate": 8.382771972506084e-06,
"loss": 0.4922539234161377,
"step": 10740
},
{
"epoch": 0.7720482619936799,
"grad_norm": 1.2196768116628702,
"learning_rate": 8.08059784488198e-06,
"loss": 0.46504602432250974,
"step": 10750
},
{
"epoch": 0.7727664464234415,
"grad_norm": 1.608227913416772,
"learning_rate": 7.780302851095544e-06,
"loss": 0.4692853927612305,
"step": 10760
},
{
"epoch": 0.7734846308532031,
"grad_norm": 0.6340817856296845,
"learning_rate": 7.482173522002166e-06,
"loss": 0.5066726207733154,
"step": 10770
},
{
"epoch": 0.7742028152829646,
"grad_norm": 0.7944404569129905,
"learning_rate": 7.186494322056624e-06,
"loss": 0.4935319900512695,
"step": 10780
},
{
"epoch": 0.7749209997127262,
"grad_norm": 0.703687204336021,
"learning_rate": 6.893547377887183e-06,
"loss": 0.47417163848876953,
"step": 10790
},
{
"epoch": 0.7756391841424878,
"grad_norm": 0.742874008812156,
"learning_rate": 6.6036122091003365e-06,
"loss": 0.44792890548706055,
"step": 10800
},
{
"epoch": 0.7763573685722494,
"grad_norm": 0.6360013796264571,
"learning_rate": 6.316965461573097e-06,
"loss": 0.4842789649963379,
"step": 10810
},
{
"epoch": 0.7770755530020109,
"grad_norm": 0.8139760009889071,
"learning_rate": 6.0338806434872965e-06,
"loss": 0.4965799331665039,
"step": 10820
},
{
"epoch": 0.7777937374317725,
"grad_norm": 0.8301886604433777,
"learning_rate": 5.754627864357749e-06,
"loss": 0.4615668296813965,
"step": 10830
},
{
"epoch": 0.7785119218615341,
"grad_norm": 0.6926304380475539,
"learning_rate": 5.479473577303319e-06,
"loss": 0.4657893180847168,
"step": 10840
},
{
"epoch": 0.7792301062912956,
"grad_norm": 0.5890313181511685,
"learning_rate": 5.208680324806778e-06,
"loss": 0.46862001419067384,
"step": 10850
},
{
"epoch": 0.7799482907210572,
"grad_norm": 0.9883696710585612,
"learning_rate": 4.94250648820606e-06,
"loss": 0.496218204498291,
"step": 10860
},
{
"epoch": 0.7806664751508188,
"grad_norm": 0.5887592389667006,
"learning_rate": 4.681206041155951e-06,
"loss": 0.4633312225341797,
"step": 10870
},
{
"epoch": 0.7813846595805803,
"grad_norm": 0.8724046827663996,
"learning_rate": 4.42502830729538e-06,
"loss": 0.4831571102142334,
"step": 10880
},
{
"epoch": 0.7821028440103418,
"grad_norm": 1.3246382009699225,
"learning_rate": 4.174217722351669e-06,
"loss": 0.48696060180664064,
"step": 10890
},
{
"epoch": 0.7828210284401034,
"grad_norm": 0.7385847674732199,
"learning_rate": 3.929013600908575e-06,
"loss": 0.4776357650756836,
"step": 10900
},
{
"epoch": 0.7835392128698649,
"grad_norm": 0.7417553280763868,
"learning_rate": 3.689649908060807e-06,
"loss": 0.4849376678466797,
"step": 10910
},
{
"epoch": 0.7842573972996265,
"grad_norm": 0.6667813686863713,
"learning_rate": 3.4563550361728028e-06,
"loss": 0.48846683502197263,
"step": 10920
},
{
"epoch": 0.7849755817293881,
"grad_norm": 0.7139720952334795,
"learning_rate": 3.2293515869548216e-06,
"loss": 0.4710516929626465,
"step": 10930
},
{
"epoch": 0.7856937661591497,
"grad_norm": 0.8091339769856745,
"learning_rate": 3.0088561590642746e-06,
"loss": 0.4894590377807617,
"step": 10940
},
{
"epoch": 0.7864119505889112,
"grad_norm": 0.817057406969214,
"learning_rate": 2.795079141434941e-06,
"loss": 0.4925840854644775,
"step": 10950
},
{
"epoch": 0.7871301350186728,
"grad_norm": 0.7412538485970538,
"learning_rate": 2.588224512531299e-06,
"loss": 0.45986323356628417,
"step": 10960
},
{
"epoch": 0.7878483194484344,
"grad_norm": 0.8206862413869445,
"learning_rate": 2.388489645719495e-06,
"loss": 0.46504931449890136,
"step": 10970
},
{
"epoch": 0.7885665038781959,
"grad_norm": 0.6553841342094755,
"learning_rate": 2.196065120940663e-06,
"loss": 0.4511693477630615,
"step": 10980
},
{
"epoch": 0.7892846883079575,
"grad_norm": 1.0803490192381104,
"learning_rate": 2.0111345428662808e-06,
"loss": 0.46965513229370115,
"step": 10990
},
{
"epoch": 0.7900028727377191,
"grad_norm": 0.7331792808350914,
"learning_rate": 1.8338743657091102e-06,
"loss": 0.48313174247741697,
"step": 11000
},
{
"epoch": 0.7907210571674806,
"grad_norm": 0.6445327459918698,
"learning_rate": 1.6644537248568113e-06,
"loss": 0.48308510780334474,
"step": 11010
},
{
"epoch": 0.7914392415972422,
"grad_norm": 0.9377684999901211,
"learning_rate": 1.503034275488951e-06,
"loss": 0.4660682678222656,
"step": 11020
},
{
"epoch": 0.7921574260270038,
"grad_norm": 0.6933315053471646,
"learning_rate": 1.3497700383313495e-06,
"loss": 0.4733739852905273,
"step": 11030
},
{
"epoch": 0.7928756104567652,
"grad_norm": 0.6806724563884676,
"learning_rate": 1.2048072526949562e-06,
"loss": 0.45346903800964355,
"step": 11040
},
{
"epoch": 0.7935937948865268,
"grad_norm": 0.6972931648423332,
"learning_rate": 1.0682842369394756e-06,
"loss": 0.4511551380157471,
"step": 11050
},
{
"epoch": 0.7943119793162884,
"grad_norm": 0.6325521976901317,
"learning_rate": 9.403312564948792e-07,
"loss": 0.4536858558654785,
"step": 11060
},
{
"epoch": 0.79503016374605,
"grad_norm": 0.6573503383463555,
"learning_rate": 8.210703995667471e-07,
"loss": 0.47881460189819336,
"step": 11070
},
{
"epoch": 0.7957483481758115,
"grad_norm": 0.6456221585258801,
"learning_rate": 7.106154606440245e-07,
"loss": 0.4543264389038086,
"step": 11080
},
{
"epoch": 0.7964665326055731,
"grad_norm": 0.6686073937576581,
"learning_rate": 6.090718319203393e-07,
"loss": 0.5069967269897461,
"step": 11090
},
{
"epoch": 0.7971847170353347,
"grad_norm": 0.780743637610021,
"learning_rate": 5.165364027325089e-07,
"loss": 0.5003859519958496,
"step": 11100
},
{
"epoch": 0.7979029014650962,
"grad_norm": 0.7094932859461007,
"learning_rate": 4.330974671121585e-07,
"loss": 0.48654727935791015,
"step": 11110
},
{
"epoch": 0.7986210858948578,
"grad_norm": 0.9474424648111287,
"learning_rate": 3.588346395386846e-07,
"loss": 0.48462328910827634,
"step": 11120
},
{
"epoch": 0.7993392703246194,
"grad_norm": 0.8410633769922748,
"learning_rate": 2.938187789739395e-07,
"loss": 0.47039108276367186,
"step": 11130
},
{
"epoch": 0.800057454754381,
"grad_norm": 0.7304868038287088,
"learning_rate": 2.3811192125110724e-07,
"loss": 0.47118682861328126,
"step": 11140
},
{
"epoch": 0.8007756391841425,
"grad_norm": 0.72390356085032,
"learning_rate": 1.917672198823177e-07,
"loss": 0.47379493713378906,
"step": 11150
},
{
"epoch": 0.8014938236139041,
"grad_norm": 0.6501978255422738,
"learning_rate": 1.5482889534143795e-07,
"loss": 0.47515525817871096,
"step": 11160
},
{
"epoch": 0.8022120080436657,
"grad_norm": 0.6632540544791249,
"learning_rate": 1.273321928704694e-07,
"loss": 0.45771117210388185,
"step": 11170
},
{
"epoch": 0.8029301924734272,
"grad_norm": 0.6851551912872883,
"learning_rate": 1.0930334884977866e-07,
"loss": 0.46851038932800293,
"step": 11180
},
{
"epoch": 0.8036483769031887,
"grad_norm": 0.7382812388709791,
"learning_rate": 1.0075956576427487e-07,
"loss": 0.47996134757995607,
"step": 11190
},
{
"epoch": 0.8038638322321172,
"step": 11193,
"total_flos": 5300110976090112.0,
"train_loss": 0.04410098469801587,
"train_runtime": 14829.0967,
"train_samples_per_second": 9.058,
"train_steps_per_second": 0.755
}
],
"logging_steps": 10,
"max_steps": 11193,
"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": 5300110976090112.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}