gpt_48 / trainer_state.json
gokulsrinivasagan's picture
End of training
1724d95 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.3934042516026481,
"eval_steps": 1000000,
"global_step": 45044,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0043668885046027,
"grad_norm": 1.662419319152832,
"learning_rate": 9.99956331114954e-06,
"loss": 3.2912,
"step": 500
},
{
"epoch": 0.0087337770092054,
"grad_norm": 1.6752963066101074,
"learning_rate": 9.99912662229908e-06,
"loss": 3.2937,
"step": 1000
},
{
"epoch": 0.013100665513808101,
"grad_norm": 1.7080398797988892,
"learning_rate": 9.99868993344862e-06,
"loss": 3.2899,
"step": 1500
},
{
"epoch": 0.0174675540184108,
"grad_norm": 1.6542549133300781,
"learning_rate": 9.99825324459816e-06,
"loss": 3.2965,
"step": 2000
},
{
"epoch": 0.021834442523013503,
"grad_norm": 1.6882691383361816,
"learning_rate": 9.9978165557477e-06,
"loss": 3.2924,
"step": 2500
},
{
"epoch": 0.026201331027616202,
"grad_norm": 1.632838249206543,
"learning_rate": 9.997379866897238e-06,
"loss": 3.2883,
"step": 3000
},
{
"epoch": 0.030568219532218905,
"grad_norm": 1.6763091087341309,
"learning_rate": 9.996943178046778e-06,
"loss": 3.2924,
"step": 3500
},
{
"epoch": 0.0349351080368216,
"grad_norm": 1.6624484062194824,
"learning_rate": 9.996506489196318e-06,
"loss": 3.2943,
"step": 4000
},
{
"epoch": 0.03930199654142431,
"grad_norm": 1.7318781614303589,
"learning_rate": 9.996069800345858e-06,
"loss": 3.2896,
"step": 4500
},
{
"epoch": 0.043668885046027006,
"grad_norm": 1.6842418909072876,
"learning_rate": 9.995633111495399e-06,
"loss": 3.2884,
"step": 5000
},
{
"epoch": 0.048035773550629705,
"grad_norm": 1.7172280550003052,
"learning_rate": 9.995196422644939e-06,
"loss": 3.2859,
"step": 5500
},
{
"epoch": 0.052402662055232405,
"grad_norm": 1.7265543937683105,
"learning_rate": 9.994759733794479e-06,
"loss": 3.2846,
"step": 6000
},
{
"epoch": 0.056769550559835104,
"grad_norm": 1.7273930311203003,
"learning_rate": 9.994323044944017e-06,
"loss": 3.2765,
"step": 6500
},
{
"epoch": 0.06113643906443781,
"grad_norm": 1.7635784149169922,
"learning_rate": 9.993886356093557e-06,
"loss": 3.2867,
"step": 7000
},
{
"epoch": 0.06550332756904051,
"grad_norm": 1.6839983463287354,
"learning_rate": 9.993449667243097e-06,
"loss": 3.2808,
"step": 7500
},
{
"epoch": 0.0698702160736432,
"grad_norm": 1.7378562688827515,
"learning_rate": 9.993012978392636e-06,
"loss": 3.2814,
"step": 8000
},
{
"epoch": 0.07423710457824591,
"grad_norm": 1.7775222063064575,
"learning_rate": 9.992576289542176e-06,
"loss": 3.2834,
"step": 8500
},
{
"epoch": 0.07860399308284861,
"grad_norm": 1.7021666765213013,
"learning_rate": 9.992139600691716e-06,
"loss": 3.2721,
"step": 9000
},
{
"epoch": 0.0829708815874513,
"grad_norm": 1.8044604063034058,
"learning_rate": 9.991702911841256e-06,
"loss": 3.2731,
"step": 9500
},
{
"epoch": 0.08733777009205401,
"grad_norm": 1.7951667308807373,
"learning_rate": 9.991266222990796e-06,
"loss": 3.2762,
"step": 10000
},
{
"epoch": 0.0917046585966567,
"grad_norm": 1.6602391004562378,
"learning_rate": 9.990829534140335e-06,
"loss": 3.2846,
"step": 10500
},
{
"epoch": 0.09607154710125941,
"grad_norm": 1.7674407958984375,
"learning_rate": 9.990392845289875e-06,
"loss": 3.2809,
"step": 11000
},
{
"epoch": 0.10043843560586212,
"grad_norm": 1.7032283544540405,
"learning_rate": 9.989956156439415e-06,
"loss": 3.2735,
"step": 11500
},
{
"epoch": 0.10480532411046481,
"grad_norm": 1.687095284461975,
"learning_rate": 9.989519467588953e-06,
"loss": 3.2595,
"step": 12000
},
{
"epoch": 0.10917221261506752,
"grad_norm": 1.6990588903427124,
"learning_rate": 9.989082778738493e-06,
"loss": 3.2719,
"step": 12500
},
{
"epoch": 0.11353910111967021,
"grad_norm": 1.7480804920196533,
"learning_rate": 9.988646089888033e-06,
"loss": 3.2718,
"step": 13000
},
{
"epoch": 0.11790598962427291,
"grad_norm": 1.6896662712097168,
"learning_rate": 9.988209401037573e-06,
"loss": 3.2716,
"step": 13500
},
{
"epoch": 0.12227287812887562,
"grad_norm": 1.6743825674057007,
"learning_rate": 9.987772712187114e-06,
"loss": 3.2636,
"step": 14000
},
{
"epoch": 0.1266397666334783,
"grad_norm": 1.72151517868042,
"learning_rate": 9.987336023336654e-06,
"loss": 3.2714,
"step": 14500
},
{
"epoch": 0.13100665513808102,
"grad_norm": 1.7318052053451538,
"learning_rate": 9.986899334486194e-06,
"loss": 3.2657,
"step": 15000
},
{
"epoch": 0.13537354364268372,
"grad_norm": 1.6856112480163574,
"learning_rate": 9.986462645635732e-06,
"loss": 3.2709,
"step": 15500
},
{
"epoch": 0.1397404321472864,
"grad_norm": 1.7183128595352173,
"learning_rate": 9.986025956785272e-06,
"loss": 3.259,
"step": 16000
},
{
"epoch": 0.1441073206518891,
"grad_norm": 1.6908427476882935,
"learning_rate": 9.985589267934812e-06,
"loss": 3.2685,
"step": 16500
},
{
"epoch": 0.14847420915649182,
"grad_norm": 1.6833419799804688,
"learning_rate": 9.985152579084352e-06,
"loss": 3.2578,
"step": 17000
},
{
"epoch": 0.15284109766109452,
"grad_norm": 1.7637497186660767,
"learning_rate": 9.984715890233891e-06,
"loss": 3.2578,
"step": 17500
},
{
"epoch": 0.15720798616569723,
"grad_norm": 1.7534202337265015,
"learning_rate": 9.984279201383431e-06,
"loss": 3.2577,
"step": 18000
},
{
"epoch": 0.1615748746702999,
"grad_norm": 1.6656574010849,
"learning_rate": 9.983842512532971e-06,
"loss": 3.2658,
"step": 18500
},
{
"epoch": 0.1659417631749026,
"grad_norm": 1.680647850036621,
"learning_rate": 9.98340582368251e-06,
"loss": 3.2593,
"step": 19000
},
{
"epoch": 0.17030865167950532,
"grad_norm": 1.6714856624603271,
"learning_rate": 9.98296913483205e-06,
"loss": 3.259,
"step": 19500
},
{
"epoch": 0.17467554018410802,
"grad_norm": 1.679284930229187,
"learning_rate": 9.98253244598159e-06,
"loss": 3.2644,
"step": 20000
},
{
"epoch": 0.17904242868871073,
"grad_norm": 1.6778266429901123,
"learning_rate": 9.982095757131129e-06,
"loss": 3.2524,
"step": 20500
},
{
"epoch": 0.1834093171933134,
"grad_norm": 1.7027274370193481,
"learning_rate": 9.981659068280668e-06,
"loss": 3.262,
"step": 21000
},
{
"epoch": 0.18777620569791612,
"grad_norm": 1.682574987411499,
"learning_rate": 9.981222379430208e-06,
"loss": 3.2549,
"step": 21500
},
{
"epoch": 0.19214309420251882,
"grad_norm": 1.656554102897644,
"learning_rate": 9.980785690579748e-06,
"loss": 3.2588,
"step": 22000
},
{
"epoch": 0.19650998270712153,
"grad_norm": 1.6965105533599854,
"learning_rate": 9.980349001729288e-06,
"loss": 3.2531,
"step": 22500
},
{
"epoch": 0.20087687121172423,
"grad_norm": 1.7153074741363525,
"learning_rate": 9.97991231287883e-06,
"loss": 3.2569,
"step": 23000
},
{
"epoch": 0.2052437597163269,
"grad_norm": 1.7081422805786133,
"learning_rate": 9.97947562402837e-06,
"loss": 3.2563,
"step": 23500
},
{
"epoch": 0.20961064822092962,
"grad_norm": 1.6499897241592407,
"learning_rate": 9.979038935177907e-06,
"loss": 3.2509,
"step": 24000
},
{
"epoch": 0.21397753672553232,
"grad_norm": 1.7635197639465332,
"learning_rate": 9.978602246327447e-06,
"loss": 3.2444,
"step": 24500
},
{
"epoch": 0.21834442523013503,
"grad_norm": 1.6655809879302979,
"learning_rate": 9.978165557476987e-06,
"loss": 3.248,
"step": 25000
},
{
"epoch": 0.22271131373473774,
"grad_norm": 1.6376826763153076,
"learning_rate": 9.977728868626527e-06,
"loss": 3.2454,
"step": 25500
},
{
"epoch": 0.22707820223934042,
"grad_norm": 1.7528952360153198,
"learning_rate": 9.977292179776067e-06,
"loss": 3.2523,
"step": 26000
},
{
"epoch": 0.23144509074394312,
"grad_norm": 1.6970986127853394,
"learning_rate": 9.976855490925607e-06,
"loss": 3.2456,
"step": 26500
},
{
"epoch": 0.23581197924854583,
"grad_norm": 1.7972602844238281,
"learning_rate": 9.976418802075146e-06,
"loss": 3.242,
"step": 27000
},
{
"epoch": 0.24017886775314853,
"grad_norm": 1.6818068027496338,
"learning_rate": 9.975982113224686e-06,
"loss": 3.2524,
"step": 27500
},
{
"epoch": 0.24454575625775124,
"grad_norm": 1.6681532859802246,
"learning_rate": 9.975545424374226e-06,
"loss": 3.2458,
"step": 28000
},
{
"epoch": 0.24891264476235392,
"grad_norm": 1.6959565877914429,
"learning_rate": 9.975108735523766e-06,
"loss": 3.2423,
"step": 28500
},
{
"epoch": 0.2532795332669566,
"grad_norm": 1.6236752271652222,
"learning_rate": 9.974672046673306e-06,
"loss": 3.2429,
"step": 29000
},
{
"epoch": 0.25764642177155933,
"grad_norm": 1.7038185596466064,
"learning_rate": 9.974235357822844e-06,
"loss": 3.2309,
"step": 29500
},
{
"epoch": 0.26201331027616204,
"grad_norm": 1.6719776391983032,
"learning_rate": 9.973798668972384e-06,
"loss": 3.243,
"step": 30000
},
{
"epoch": 0.26638019878076474,
"grad_norm": 1.6746853590011597,
"learning_rate": 9.973361980121923e-06,
"loss": 3.2439,
"step": 30500
},
{
"epoch": 0.27074708728536745,
"grad_norm": 1.755371332168579,
"learning_rate": 9.972925291271463e-06,
"loss": 3.2387,
"step": 31000
},
{
"epoch": 0.27511397578997016,
"grad_norm": 1.7034052610397339,
"learning_rate": 9.972488602421003e-06,
"loss": 3.2348,
"step": 31500
},
{
"epoch": 0.2794808642945728,
"grad_norm": 1.6981998682022095,
"learning_rate": 9.972051913570545e-06,
"loss": 3.2316,
"step": 32000
},
{
"epoch": 0.2838477527991755,
"grad_norm": 1.6613941192626953,
"learning_rate": 9.971615224720084e-06,
"loss": 3.2353,
"step": 32500
},
{
"epoch": 0.2882146413037782,
"grad_norm": 1.6410475969314575,
"learning_rate": 9.971178535869623e-06,
"loss": 3.2273,
"step": 33000
},
{
"epoch": 0.2925815298083809,
"grad_norm": 1.6354302167892456,
"learning_rate": 9.970741847019162e-06,
"loss": 3.2294,
"step": 33500
},
{
"epoch": 0.29694841831298363,
"grad_norm": 1.7013846635818481,
"learning_rate": 9.970305158168702e-06,
"loss": 3.2283,
"step": 34000
},
{
"epoch": 0.30131530681758634,
"grad_norm": 1.66492760181427,
"learning_rate": 9.969868469318242e-06,
"loss": 3.236,
"step": 34500
},
{
"epoch": 0.30568219532218904,
"grad_norm": 1.7220481634140015,
"learning_rate": 9.969431780467782e-06,
"loss": 3.2312,
"step": 35000
},
{
"epoch": 0.31004908382679175,
"grad_norm": 1.705837368965149,
"learning_rate": 9.968995091617322e-06,
"loss": 3.2259,
"step": 35500
},
{
"epoch": 0.31441597233139446,
"grad_norm": 1.6827079057693481,
"learning_rate": 9.968558402766862e-06,
"loss": 3.2251,
"step": 36000
},
{
"epoch": 0.31878286083599716,
"grad_norm": 1.7346361875534058,
"learning_rate": 9.968121713916401e-06,
"loss": 3.2243,
"step": 36500
},
{
"epoch": 0.3231497493405998,
"grad_norm": 1.7183014154434204,
"learning_rate": 9.967685025065941e-06,
"loss": 3.2248,
"step": 37000
},
{
"epoch": 0.3275166378452025,
"grad_norm": 1.6337991952896118,
"learning_rate": 9.967248336215481e-06,
"loss": 3.2277,
"step": 37500
},
{
"epoch": 0.3318835263498052,
"grad_norm": 1.6129624843597412,
"learning_rate": 9.96681164736502e-06,
"loss": 3.2251,
"step": 38000
},
{
"epoch": 0.33625041485440793,
"grad_norm": 1.6240402460098267,
"learning_rate": 9.966374958514559e-06,
"loss": 3.2271,
"step": 38500
},
{
"epoch": 0.34061730335901064,
"grad_norm": 1.6896299123764038,
"learning_rate": 9.965938269664099e-06,
"loss": 3.2229,
"step": 39000
},
{
"epoch": 0.34498419186361334,
"grad_norm": 1.6979056596755981,
"learning_rate": 9.965501580813639e-06,
"loss": 3.2242,
"step": 39500
},
{
"epoch": 0.34935108036821605,
"grad_norm": 1.6829845905303955,
"learning_rate": 9.965064891963178e-06,
"loss": 3.2189,
"step": 40000
},
{
"epoch": 0.35371796887281876,
"grad_norm": 1.6688014268875122,
"learning_rate": 9.964628203112718e-06,
"loss": 3.2194,
"step": 40500
},
{
"epoch": 0.35808485737742146,
"grad_norm": 1.6810382604599,
"learning_rate": 9.96419151426226e-06,
"loss": 3.2226,
"step": 41000
},
{
"epoch": 0.36245174588202417,
"grad_norm": 1.6257010698318481,
"learning_rate": 9.9637548254118e-06,
"loss": 3.2177,
"step": 41500
},
{
"epoch": 0.3668186343866268,
"grad_norm": 1.6654291152954102,
"learning_rate": 9.963318136561338e-06,
"loss": 3.2188,
"step": 42000
},
{
"epoch": 0.3711855228912295,
"grad_norm": 1.622125506401062,
"learning_rate": 9.962881447710878e-06,
"loss": 3.2136,
"step": 42500
},
{
"epoch": 0.37555241139583223,
"grad_norm": 1.6622231006622314,
"learning_rate": 9.962444758860417e-06,
"loss": 3.2169,
"step": 43000
},
{
"epoch": 0.37991929990043494,
"grad_norm": 1.7069813013076782,
"learning_rate": 9.962008070009957e-06,
"loss": 3.2189,
"step": 43500
},
{
"epoch": 0.38428618840503764,
"grad_norm": 1.6623612642288208,
"learning_rate": 9.961571381159497e-06,
"loss": 3.2123,
"step": 44000
},
{
"epoch": 0.38865307690964035,
"grad_norm": 1.6388291120529175,
"learning_rate": 9.961134692309037e-06,
"loss": 3.2068,
"step": 44500
},
{
"epoch": 0.39301996541424306,
"grad_norm": 1.6568938493728638,
"learning_rate": 9.960698003458577e-06,
"loss": 3.2155,
"step": 45000
},
{
"epoch": 0.3934042516026481,
"step": 45044,
"total_flos": 1.1298852697433702e+18,
"train_loss": 3.2530452862756416,
"train_runtime": 46795.5118,
"train_samples_per_second": 11744.43,
"train_steps_per_second": 244.677
}
],
"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": 1.1298852697433702e+18,
"train_batch_size": 48,
"trial_name": null,
"trial_params": null
}