gpt_24 / trainer_state.json
gokulsrinivasagan's picture
End of training
cb7632e verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.4035965693723908,
"eval_steps": 1000000,
"global_step": 46211,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0043668885046027,
"grad_norm": 1.8038700819015503,
"learning_rate": 9.99956331114954e-06,
"loss": 3.7275,
"step": 500
},
{
"epoch": 0.0087337770092054,
"grad_norm": 1.882214069366455,
"learning_rate": 9.99912662229908e-06,
"loss": 3.7247,
"step": 1000
},
{
"epoch": 0.013100665513808101,
"grad_norm": 1.934370994567871,
"learning_rate": 9.99868993344862e-06,
"loss": 3.7176,
"step": 1500
},
{
"epoch": 0.0174675540184108,
"grad_norm": 1.8633307218551636,
"learning_rate": 9.99825324459816e-06,
"loss": 3.7196,
"step": 2000
},
{
"epoch": 0.021834442523013503,
"grad_norm": 1.8951599597930908,
"learning_rate": 9.9978165557477e-06,
"loss": 3.7127,
"step": 2500
},
{
"epoch": 0.026201331027616202,
"grad_norm": 1.875725507736206,
"learning_rate": 9.997379866897238e-06,
"loss": 3.7033,
"step": 3000
},
{
"epoch": 0.030568219532218905,
"grad_norm": 1.8393280506134033,
"learning_rate": 9.996943178046778e-06,
"loss": 3.7037,
"step": 3500
},
{
"epoch": 0.0349351080368216,
"grad_norm": 2.0136420726776123,
"learning_rate": 9.996506489196318e-06,
"loss": 3.7021,
"step": 4000
},
{
"epoch": 0.03930199654142431,
"grad_norm": 1.8446974754333496,
"learning_rate": 9.996069800345858e-06,
"loss": 3.6928,
"step": 4500
},
{
"epoch": 0.043668885046027006,
"grad_norm": 1.8116663694381714,
"learning_rate": 9.995633111495399e-06,
"loss": 3.6877,
"step": 5000
},
{
"epoch": 0.048035773550629705,
"grad_norm": 1.8685935735702515,
"learning_rate": 9.995196422644939e-06,
"loss": 3.6815,
"step": 5500
},
{
"epoch": 0.052402662055232405,
"grad_norm": 1.8733059167861938,
"learning_rate": 9.994759733794479e-06,
"loss": 3.6753,
"step": 6000
},
{
"epoch": 0.056769550559835104,
"grad_norm": 1.9478683471679688,
"learning_rate": 9.994323044944017e-06,
"loss": 3.6652,
"step": 6500
},
{
"epoch": 0.06113643906443781,
"grad_norm": 1.9699769020080566,
"learning_rate": 9.993886356093557e-06,
"loss": 3.6716,
"step": 7000
},
{
"epoch": 0.06550332756904051,
"grad_norm": 1.8126554489135742,
"learning_rate": 9.993449667243097e-06,
"loss": 3.6619,
"step": 7500
},
{
"epoch": 0.0698702160736432,
"grad_norm": 1.965243935585022,
"learning_rate": 9.993012978392636e-06,
"loss": 3.6587,
"step": 8000
},
{
"epoch": 0.07423710457824591,
"grad_norm": 1.8856313228607178,
"learning_rate": 9.992576289542176e-06,
"loss": 3.6573,
"step": 8500
},
{
"epoch": 0.07860399308284861,
"grad_norm": 1.9191926717758179,
"learning_rate": 9.992139600691716e-06,
"loss": 3.6421,
"step": 9000
},
{
"epoch": 0.0829708815874513,
"grad_norm": 1.9937667846679688,
"learning_rate": 9.991702911841256e-06,
"loss": 3.6399,
"step": 9500
},
{
"epoch": 0.08733777009205401,
"grad_norm": 1.911783218383789,
"learning_rate": 9.991266222990796e-06,
"loss": 3.641,
"step": 10000
},
{
"epoch": 0.0917046585966567,
"grad_norm": 1.9108695983886719,
"learning_rate": 9.990829534140335e-06,
"loss": 3.6448,
"step": 10500
},
{
"epoch": 0.09607154710125941,
"grad_norm": 1.9195294380187988,
"learning_rate": 9.990392845289875e-06,
"loss": 3.6369,
"step": 11000
},
{
"epoch": 0.10043843560586212,
"grad_norm": 1.8656957149505615,
"learning_rate": 9.989956156439415e-06,
"loss": 3.6278,
"step": 11500
},
{
"epoch": 0.10480532411046481,
"grad_norm": 1.8072896003723145,
"learning_rate": 9.989519467588953e-06,
"loss": 3.6106,
"step": 12000
},
{
"epoch": 0.10917221261506752,
"grad_norm": 1.8502795696258545,
"learning_rate": 9.989082778738493e-06,
"loss": 3.6208,
"step": 12500
},
{
"epoch": 0.11353910111967021,
"grad_norm": 1.9243334531784058,
"learning_rate": 9.988646089888033e-06,
"loss": 3.6175,
"step": 13000
},
{
"epoch": 0.11790598962427291,
"grad_norm": 1.8611966371536255,
"learning_rate": 9.988209401037573e-06,
"loss": 3.6144,
"step": 13500
},
{
"epoch": 0.12227287812887562,
"grad_norm": 1.8360419273376465,
"learning_rate": 9.987772712187114e-06,
"loss": 3.6033,
"step": 14000
},
{
"epoch": 0.1266397666334783,
"grad_norm": 1.9218759536743164,
"learning_rate": 9.987336023336654e-06,
"loss": 3.6078,
"step": 14500
},
{
"epoch": 0.13100665513808102,
"grad_norm": 1.868091344833374,
"learning_rate": 9.986899334486194e-06,
"loss": 3.6002,
"step": 15000
},
{
"epoch": 0.13537354364268372,
"grad_norm": 1.8368401527404785,
"learning_rate": 9.986462645635732e-06,
"loss": 3.6028,
"step": 15500
},
{
"epoch": 0.1397404321472864,
"grad_norm": 1.838019847869873,
"learning_rate": 9.986025956785272e-06,
"loss": 3.5879,
"step": 16000
},
{
"epoch": 0.1441073206518891,
"grad_norm": 1.816603422164917,
"learning_rate": 9.985589267934812e-06,
"loss": 3.5938,
"step": 16500
},
{
"epoch": 0.14847420915649182,
"grad_norm": 1.8263438940048218,
"learning_rate": 9.985152579084352e-06,
"loss": 3.5808,
"step": 17000
},
{
"epoch": 0.15284109766109452,
"grad_norm": 1.8684626817703247,
"learning_rate": 9.984715890233891e-06,
"loss": 3.578,
"step": 17500
},
{
"epoch": 0.15720798616569723,
"grad_norm": 1.9385733604431152,
"learning_rate": 9.984279201383431e-06,
"loss": 3.5759,
"step": 18000
},
{
"epoch": 0.1615748746702999,
"grad_norm": 1.8138542175292969,
"learning_rate": 9.983842512532971e-06,
"loss": 3.5804,
"step": 18500
},
{
"epoch": 0.1659417631749026,
"grad_norm": 1.8231014013290405,
"learning_rate": 9.98340582368251e-06,
"loss": 3.5728,
"step": 19000
},
{
"epoch": 0.17030865167950532,
"grad_norm": 1.820784568786621,
"learning_rate": 9.98296913483205e-06,
"loss": 3.5699,
"step": 19500
},
{
"epoch": 0.17467554018410802,
"grad_norm": 1.8315212726593018,
"learning_rate": 9.98253244598159e-06,
"loss": 3.5727,
"step": 20000
},
{
"epoch": 0.17904242868871073,
"grad_norm": 1.8523602485656738,
"learning_rate": 9.982095757131129e-06,
"loss": 3.5587,
"step": 20500
},
{
"epoch": 0.1834093171933134,
"grad_norm": 1.8133721351623535,
"learning_rate": 9.981659068280668e-06,
"loss": 3.5656,
"step": 21000
},
{
"epoch": 0.18777620569791612,
"grad_norm": 1.826695203781128,
"learning_rate": 9.981222379430208e-06,
"loss": 3.5574,
"step": 21500
},
{
"epoch": 0.19214309420251882,
"grad_norm": 1.7944340705871582,
"learning_rate": 9.980785690579748e-06,
"loss": 3.5574,
"step": 22000
},
{
"epoch": 0.19650998270712153,
"grad_norm": 1.8452637195587158,
"learning_rate": 9.980349001729288e-06,
"loss": 3.5502,
"step": 22500
},
{
"epoch": 0.20087687121172423,
"grad_norm": 1.8428194522857666,
"learning_rate": 9.97991231287883e-06,
"loss": 3.5518,
"step": 23000
},
{
"epoch": 0.2052437597163269,
"grad_norm": 1.873677134513855,
"learning_rate": 9.97947562402837e-06,
"loss": 3.5488,
"step": 23500
},
{
"epoch": 0.20961064822092962,
"grad_norm": 1.819214940071106,
"learning_rate": 9.979038935177907e-06,
"loss": 3.5398,
"step": 24000
},
{
"epoch": 0.21397753672553232,
"grad_norm": 1.8920146226882935,
"learning_rate": 9.978602246327447e-06,
"loss": 3.5324,
"step": 24500
},
{
"epoch": 0.21834442523013503,
"grad_norm": 1.831764817237854,
"learning_rate": 9.978165557476987e-06,
"loss": 3.5342,
"step": 25000
},
{
"epoch": 0.22271131373473774,
"grad_norm": 1.8074841499328613,
"learning_rate": 9.977728868626527e-06,
"loss": 3.53,
"step": 25500
},
{
"epoch": 0.22707820223934042,
"grad_norm": 1.9136210680007935,
"learning_rate": 9.977292179776067e-06,
"loss": 3.5342,
"step": 26000
},
{
"epoch": 0.23144509074394312,
"grad_norm": 1.8257259130477905,
"learning_rate": 9.976855490925607e-06,
"loss": 3.5245,
"step": 26500
},
{
"epoch": 0.23581197924854583,
"grad_norm": 1.9069098234176636,
"learning_rate": 9.976418802075146e-06,
"loss": 3.5204,
"step": 27000
},
{
"epoch": 0.24017886775314853,
"grad_norm": 1.7865424156188965,
"learning_rate": 9.975982113224686e-06,
"loss": 3.5283,
"step": 27500
},
{
"epoch": 0.24454575625775124,
"grad_norm": 1.8387185335159302,
"learning_rate": 9.975545424374226e-06,
"loss": 3.5199,
"step": 28000
},
{
"epoch": 0.24891264476235392,
"grad_norm": 1.8645451068878174,
"learning_rate": 9.975108735523766e-06,
"loss": 3.5148,
"step": 28500
},
{
"epoch": 0.2532795332669566,
"grad_norm": 1.7989400625228882,
"learning_rate": 9.974672046673306e-06,
"loss": 3.5133,
"step": 29000
},
{
"epoch": 0.25764642177155933,
"grad_norm": 1.7804815769195557,
"learning_rate": 9.974235357822844e-06,
"loss": 3.4984,
"step": 29500
},
{
"epoch": 0.26201331027616204,
"grad_norm": 1.844984531402588,
"learning_rate": 9.973798668972384e-06,
"loss": 3.5089,
"step": 30000
},
{
"epoch": 0.26638019878076474,
"grad_norm": 1.8139082193374634,
"learning_rate": 9.973361980121923e-06,
"loss": 3.5089,
"step": 30500
},
{
"epoch": 0.27074708728536745,
"grad_norm": 1.8752747774124146,
"learning_rate": 9.972925291271463e-06,
"loss": 3.5011,
"step": 31000
},
{
"epoch": 0.27511397578997016,
"grad_norm": 1.7818485498428345,
"learning_rate": 9.972488602421003e-06,
"loss": 3.4952,
"step": 31500
},
{
"epoch": 0.2794808642945728,
"grad_norm": 1.87028169631958,
"learning_rate": 9.972051913570545e-06,
"loss": 3.4904,
"step": 32000
},
{
"epoch": 0.2838477527991755,
"grad_norm": 1.7847636938095093,
"learning_rate": 9.971615224720084e-06,
"loss": 3.4925,
"step": 32500
},
{
"epoch": 0.2882146413037782,
"grad_norm": 1.7927122116088867,
"learning_rate": 9.971178535869623e-06,
"loss": 3.4823,
"step": 33000
},
{
"epoch": 0.2925815298083809,
"grad_norm": 1.7564101219177246,
"learning_rate": 9.970741847019162e-06,
"loss": 3.483,
"step": 33500
},
{
"epoch": 0.29694841831298363,
"grad_norm": 1.8266777992248535,
"learning_rate": 9.970305158168702e-06,
"loss": 3.4805,
"step": 34000
},
{
"epoch": 0.30131530681758634,
"grad_norm": 1.7539730072021484,
"learning_rate": 9.969868469318242e-06,
"loss": 3.4876,
"step": 34500
},
{
"epoch": 0.30568219532218904,
"grad_norm": 1.838841199874878,
"learning_rate": 9.969431780467782e-06,
"loss": 3.4802,
"step": 35000
},
{
"epoch": 0.31004908382679175,
"grad_norm": 1.7870306968688965,
"learning_rate": 9.968995091617322e-06,
"loss": 3.4738,
"step": 35500
},
{
"epoch": 0.31441597233139446,
"grad_norm": 1.8143889904022217,
"learning_rate": 9.968558402766862e-06,
"loss": 3.4714,
"step": 36000
},
{
"epoch": 0.31878286083599716,
"grad_norm": 1.9000575542449951,
"learning_rate": 9.968121713916401e-06,
"loss": 3.4694,
"step": 36500
},
{
"epoch": 0.3231497493405998,
"grad_norm": 1.8620721101760864,
"learning_rate": 9.967685025065941e-06,
"loss": 3.4683,
"step": 37000
},
{
"epoch": 0.3275166378452025,
"grad_norm": 1.7763477563858032,
"learning_rate": 9.967248336215481e-06,
"loss": 3.4687,
"step": 37500
},
{
"epoch": 0.3318835263498052,
"grad_norm": 1.730004906654358,
"learning_rate": 9.96681164736502e-06,
"loss": 3.4653,
"step": 38000
},
{
"epoch": 0.33625041485440793,
"grad_norm": 1.7360706329345703,
"learning_rate": 9.966374958514559e-06,
"loss": 3.4659,
"step": 38500
},
{
"epoch": 0.34061730335901064,
"grad_norm": 1.7843414545059204,
"learning_rate": 9.965938269664099e-06,
"loss": 3.4596,
"step": 39000
},
{
"epoch": 0.34498419186361334,
"grad_norm": 1.7715613842010498,
"learning_rate": 9.965501580813639e-06,
"loss": 3.4598,
"step": 39500
},
{
"epoch": 0.34935108036821605,
"grad_norm": 1.8175365924835205,
"learning_rate": 9.965064891963178e-06,
"loss": 3.4525,
"step": 40000
},
{
"epoch": 0.35371796887281876,
"grad_norm": 1.783474326133728,
"learning_rate": 9.964628203112718e-06,
"loss": 3.4526,
"step": 40500
},
{
"epoch": 0.35808485737742146,
"grad_norm": 1.8786745071411133,
"learning_rate": 9.96419151426226e-06,
"loss": 3.4535,
"step": 41000
},
{
"epoch": 0.36245174588202417,
"grad_norm": 1.7226436138153076,
"learning_rate": 9.9637548254118e-06,
"loss": 3.4476,
"step": 41500
},
{
"epoch": 0.3668186343866268,
"grad_norm": 1.805066704750061,
"learning_rate": 9.963318136561338e-06,
"loss": 3.4484,
"step": 42000
},
{
"epoch": 0.3711855228912295,
"grad_norm": 1.737598180770874,
"learning_rate": 9.962881447710878e-06,
"loss": 3.4411,
"step": 42500
},
{
"epoch": 0.37555241139583223,
"grad_norm": 1.7667919397354126,
"learning_rate": 9.962444758860417e-06,
"loss": 3.4428,
"step": 43000
},
{
"epoch": 0.37991929990043494,
"grad_norm": 1.8246058225631714,
"learning_rate": 9.962008070009957e-06,
"loss": 3.4436,
"step": 43500
},
{
"epoch": 0.38428618840503764,
"grad_norm": 1.8644582033157349,
"learning_rate": 9.961571381159497e-06,
"loss": 3.4355,
"step": 44000
},
{
"epoch": 0.38865307690964035,
"grad_norm": 1.7610771656036377,
"learning_rate": 9.961134692309037e-06,
"loss": 3.4286,
"step": 44500
},
{
"epoch": 0.39301996541424306,
"grad_norm": 1.7471799850463867,
"learning_rate": 9.960698003458577e-06,
"loss": 3.4359,
"step": 45000
},
{
"epoch": 0.39738685391884576,
"grad_norm": 1.730660319328308,
"learning_rate": 9.960261314608117e-06,
"loss": 3.4329,
"step": 45500
},
{
"epoch": 0.40175374242344847,
"grad_norm": 1.7533937692642212,
"learning_rate": 9.959824625757656e-06,
"loss": 3.4346,
"step": 46000
},
{
"epoch": 0.4035965693723908,
"step": 46211,
"total_flos": 1.1591583382865183e+18,
"train_loss": 3.5562555452165068,
"train_runtime": 46797.4598,
"train_samples_per_second": 11743.941,
"train_steps_per_second": 244.667
}
],
"logging_steps": 500,
"max_steps": 11449800,
"num_input_tokens_seen": 0,
"num_train_epochs": 100,
"save_steps": 1000000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1.1591583382865183e+18,
"train_batch_size": 48,
"trial_name": null,
"trial_params": null
}