alex-gpt2000 / trainer_state.json
danurahul's picture
Initial commit
9544a16
Raw
History Blame
15.1 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.9921127040031748,
"global_step": 60000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 1.994488262755538e-05,
"loss": 3.0247,
"step": 500
},
{
"epoch": 0.02,
"learning_rate": 1.988976525511076e-05,
"loss": 3.0232,
"step": 1000
},
{
"epoch": 0.02,
"learning_rate": 1.983464788266614e-05,
"loss": 3.0873,
"step": 1500
},
{
"epoch": 0.03,
"learning_rate": 1.977953051022152e-05,
"loss": 3.0891,
"step": 2000
},
{
"epoch": 0.04,
"learning_rate": 1.9724413137776898e-05,
"loss": 3.0793,
"step": 2500
},
{
"epoch": 0.05,
"learning_rate": 1.9669295765332277e-05,
"loss": 3.0651,
"step": 3000
},
{
"epoch": 0.06,
"learning_rate": 1.9614178392887656e-05,
"loss": 3.0738,
"step": 3500
},
{
"epoch": 0.07,
"learning_rate": 1.9559061020443036e-05,
"loss": 3.0549,
"step": 4000
},
{
"epoch": 0.07,
"learning_rate": 1.9503943647998415e-05,
"loss": 3.0288,
"step": 4500
},
{
"epoch": 0.08,
"learning_rate": 1.9448826275553794e-05,
"loss": 3.0467,
"step": 5000
},
{
"epoch": 0.09,
"learning_rate": 1.9393708903109173e-05,
"loss": 3.0283,
"step": 5500
},
{
"epoch": 0.1,
"learning_rate": 1.9338591530664553e-05,
"loss": 3.0267,
"step": 6000
},
{
"epoch": 0.11,
"learning_rate": 1.9283474158219932e-05,
"loss": 3.0254,
"step": 6500
},
{
"epoch": 0.12,
"learning_rate": 1.922835678577531e-05,
"loss": 3.0185,
"step": 7000
},
{
"epoch": 0.12,
"learning_rate": 1.917323941333069e-05,
"loss": 2.9999,
"step": 7500
},
{
"epoch": 0.13,
"learning_rate": 1.911812204088607e-05,
"loss": 3.0031,
"step": 8000
},
{
"epoch": 0.14,
"learning_rate": 1.9063004668441445e-05,
"loss": 3.0106,
"step": 8500
},
{
"epoch": 0.15,
"learning_rate": 1.9007887295996828e-05,
"loss": 3.0312,
"step": 9000
},
{
"epoch": 0.16,
"learning_rate": 1.8952769923552204e-05,
"loss": 3.0034,
"step": 9500
},
{
"epoch": 0.17,
"learning_rate": 1.8897652551107587e-05,
"loss": 3.0157,
"step": 10000
},
{
"epoch": 0.17,
"learning_rate": 1.8842535178662962e-05,
"loss": 2.9945,
"step": 10500
},
{
"epoch": 0.18,
"learning_rate": 1.8787417806218345e-05,
"loss": 2.9995,
"step": 11000
},
{
"epoch": 0.19,
"learning_rate": 1.873230043377372e-05,
"loss": 2.997,
"step": 11500
},
{
"epoch": 0.2,
"learning_rate": 1.8677183061329104e-05,
"loss": 3.0017,
"step": 12000
},
{
"epoch": 0.21,
"learning_rate": 1.862206568888448e-05,
"loss": 3.0046,
"step": 12500
},
{
"epoch": 0.21,
"learning_rate": 1.8566948316439862e-05,
"loss": 2.9975,
"step": 13000
},
{
"epoch": 0.22,
"learning_rate": 1.8511830943995238e-05,
"loss": 2.9778,
"step": 13500
},
{
"epoch": 0.23,
"learning_rate": 1.845671357155062e-05,
"loss": 2.9802,
"step": 14000
},
{
"epoch": 0.24,
"learning_rate": 1.8401596199105996e-05,
"loss": 2.9603,
"step": 14500
},
{
"epoch": 0.25,
"learning_rate": 1.8346478826661376e-05,
"loss": 2.9738,
"step": 15000
},
{
"epoch": 0.26,
"learning_rate": 1.8291361454216755e-05,
"loss": 2.9776,
"step": 15500
},
{
"epoch": 0.26,
"learning_rate": 1.8236244081772134e-05,
"loss": 3.0002,
"step": 16000
},
{
"epoch": 0.27,
"learning_rate": 1.8181126709327513e-05,
"loss": 3.0026,
"step": 16500
},
{
"epoch": 0.28,
"learning_rate": 1.8126009336882893e-05,
"loss": 2.9791,
"step": 17000
},
{
"epoch": 0.29,
"learning_rate": 1.8070891964438272e-05,
"loss": 2.9974,
"step": 17500
},
{
"epoch": 0.3,
"learning_rate": 1.801577459199365e-05,
"loss": 2.9739,
"step": 18000
},
{
"epoch": 0.31,
"learning_rate": 1.796065721954903e-05,
"loss": 2.9733,
"step": 18500
},
{
"epoch": 0.31,
"learning_rate": 1.790553984710441e-05,
"loss": 2.9777,
"step": 19000
},
{
"epoch": 0.32,
"learning_rate": 1.785042247465979e-05,
"loss": 2.9704,
"step": 19500
},
{
"epoch": 0.33,
"learning_rate": 1.7795305102215168e-05,
"loss": 3.0007,
"step": 20000
},
{
"epoch": 0.34,
"learning_rate": 1.7740187729770547e-05,
"loss": 2.9752,
"step": 20500
},
{
"epoch": 0.35,
"learning_rate": 1.7685070357325927e-05,
"loss": 2.9675,
"step": 21000
},
{
"epoch": 0.36,
"learning_rate": 1.7629952984881306e-05,
"loss": 2.9655,
"step": 21500
},
{
"epoch": 0.36,
"learning_rate": 1.7574835612436685e-05,
"loss": 2.9562,
"step": 22000
},
{
"epoch": 0.37,
"learning_rate": 1.7519718239992064e-05,
"loss": 2.9767,
"step": 22500
},
{
"epoch": 0.38,
"learning_rate": 1.7464600867547444e-05,
"loss": 2.9622,
"step": 23000
},
{
"epoch": 0.39,
"learning_rate": 1.7409483495102823e-05,
"loss": 2.9772,
"step": 23500
},
{
"epoch": 0.4,
"learning_rate": 1.7354366122658202e-05,
"loss": 2.947,
"step": 24000
},
{
"epoch": 0.41,
"learning_rate": 1.729924875021358e-05,
"loss": 2.9434,
"step": 24500
},
{
"epoch": 0.41,
"learning_rate": 1.724413137776896e-05,
"loss": 2.9474,
"step": 25000
},
{
"epoch": 0.42,
"learning_rate": 1.718901400532434e-05,
"loss": 2.9502,
"step": 25500
},
{
"epoch": 0.43,
"learning_rate": 1.713389663287972e-05,
"loss": 2.9575,
"step": 26000
},
{
"epoch": 0.44,
"learning_rate": 1.70787792604351e-05,
"loss": 2.9585,
"step": 26500
},
{
"epoch": 0.45,
"learning_rate": 1.7023661887990478e-05,
"loss": 2.9681,
"step": 27000
},
{
"epoch": 0.45,
"learning_rate": 1.6968544515545857e-05,
"loss": 2.9654,
"step": 27500
},
{
"epoch": 0.46,
"learning_rate": 1.6913427143101236e-05,
"loss": 2.9604,
"step": 28000
},
{
"epoch": 0.47,
"learning_rate": 1.6858309770656615e-05,
"loss": 2.9787,
"step": 28500
},
{
"epoch": 0.48,
"learning_rate": 1.6803192398211995e-05,
"loss": 2.9753,
"step": 29000
},
{
"epoch": 0.49,
"learning_rate": 1.6748075025767374e-05,
"loss": 2.9553,
"step": 29500
},
{
"epoch": 0.5,
"learning_rate": 1.6692957653322753e-05,
"loss": 2.9643,
"step": 30000
},
{
"epoch": 0.5,
"learning_rate": 1.663784028087813e-05,
"loss": 2.9379,
"step": 30500
},
{
"epoch": 0.51,
"learning_rate": 1.658272290843351e-05,
"loss": 2.9358,
"step": 31000
},
{
"epoch": 0.52,
"learning_rate": 1.6527605535988887e-05,
"loss": 2.9484,
"step": 31500
},
{
"epoch": 0.53,
"learning_rate": 1.647248816354427e-05,
"loss": 2.925,
"step": 32000
},
{
"epoch": 0.54,
"learning_rate": 1.6417370791099646e-05,
"loss": 2.9875,
"step": 32500
},
{
"epoch": 0.55,
"learning_rate": 1.636225341865503e-05,
"loss": 2.9575,
"step": 33000
},
{
"epoch": 0.55,
"learning_rate": 1.6307136046210404e-05,
"loss": 2.9438,
"step": 33500
},
{
"epoch": 0.56,
"learning_rate": 1.6252018673765787e-05,
"loss": 2.9512,
"step": 34000
},
{
"epoch": 0.57,
"learning_rate": 1.6196901301321163e-05,
"loss": 2.9298,
"step": 34500
},
{
"epoch": 0.58,
"learning_rate": 1.6141783928876546e-05,
"loss": 2.9455,
"step": 35000
},
{
"epoch": 0.59,
"learning_rate": 1.608666655643192e-05,
"loss": 2.9515,
"step": 35500
},
{
"epoch": 0.6,
"learning_rate": 1.6031549183987304e-05,
"loss": 2.9507,
"step": 36000
},
{
"epoch": 0.6,
"learning_rate": 1.597643181154268e-05,
"loss": 2.9524,
"step": 36500
},
{
"epoch": 0.61,
"learning_rate": 1.5921314439098063e-05,
"loss": 2.9466,
"step": 37000
},
{
"epoch": 0.62,
"learning_rate": 1.586619706665344e-05,
"loss": 2.9244,
"step": 37500
},
{
"epoch": 0.63,
"learning_rate": 1.581107969420882e-05,
"loss": 2.9416,
"step": 38000
},
{
"epoch": 0.64,
"learning_rate": 1.5755962321764197e-05,
"loss": 2.9553,
"step": 38500
},
{
"epoch": 0.64,
"learning_rate": 1.570084494931958e-05,
"loss": 2.9336,
"step": 39000
},
{
"epoch": 0.65,
"learning_rate": 1.5645727576874955e-05,
"loss": 2.9226,
"step": 39500
},
{
"epoch": 0.66,
"learning_rate": 1.5590610204430338e-05,
"loss": 2.9215,
"step": 40000
},
{
"epoch": 0.67,
"learning_rate": 1.5535492831985714e-05,
"loss": 2.9129,
"step": 40500
},
{
"epoch": 0.68,
"learning_rate": 1.5480375459541097e-05,
"loss": 2.9252,
"step": 41000
},
{
"epoch": 0.69,
"learning_rate": 1.5425258087096472e-05,
"loss": 2.9433,
"step": 41500
},
{
"epoch": 0.69,
"learning_rate": 1.537014071465185e-05,
"loss": 2.9279,
"step": 42000
},
{
"epoch": 0.7,
"learning_rate": 1.531502334220723e-05,
"loss": 2.9374,
"step": 42500
},
{
"epoch": 0.71,
"learning_rate": 1.525990596976261e-05,
"loss": 2.9336,
"step": 43000
},
{
"epoch": 0.72,
"learning_rate": 1.520478859731799e-05,
"loss": 2.8991,
"step": 43500
},
{
"epoch": 0.73,
"learning_rate": 1.514967122487337e-05,
"loss": 2.9326,
"step": 44000
},
{
"epoch": 0.74,
"learning_rate": 1.5094553852428748e-05,
"loss": 2.9358,
"step": 44500
},
{
"epoch": 0.74,
"learning_rate": 1.5039436479984127e-05,
"loss": 2.9192,
"step": 45000
},
{
"epoch": 0.75,
"learning_rate": 1.4984319107539506e-05,
"loss": 2.9438,
"step": 45500
},
{
"epoch": 0.76,
"learning_rate": 1.4929201735094884e-05,
"loss": 2.9213,
"step": 46000
},
{
"epoch": 0.77,
"learning_rate": 1.4874084362650265e-05,
"loss": 2.9273,
"step": 46500
},
{
"epoch": 0.78,
"learning_rate": 1.4818966990205642e-05,
"loss": 2.925,
"step": 47000
},
{
"epoch": 0.79,
"learning_rate": 1.4763849617761023e-05,
"loss": 2.9151,
"step": 47500
},
{
"epoch": 0.79,
"learning_rate": 1.4708732245316401e-05,
"loss": 2.9195,
"step": 48000
},
{
"epoch": 0.8,
"learning_rate": 1.4653614872871782e-05,
"loss": 2.9059,
"step": 48500
},
{
"epoch": 0.81,
"learning_rate": 1.459849750042716e-05,
"loss": 2.9001,
"step": 49000
},
{
"epoch": 0.82,
"learning_rate": 1.454338012798254e-05,
"loss": 2.8995,
"step": 49500
},
{
"epoch": 0.83,
"learning_rate": 1.4488262755537918e-05,
"loss": 2.9395,
"step": 50000
},
{
"epoch": 0.84,
"learning_rate": 1.4433145383093299e-05,
"loss": 2.9066,
"step": 50500
},
{
"epoch": 0.84,
"learning_rate": 1.4378028010648676e-05,
"loss": 2.8997,
"step": 51000
},
{
"epoch": 0.85,
"learning_rate": 1.4322910638204057e-05,
"loss": 2.9083,
"step": 51500
},
{
"epoch": 0.86,
"learning_rate": 1.4267793265759435e-05,
"loss": 2.9211,
"step": 52000
},
{
"epoch": 0.87,
"learning_rate": 1.4212675893314816e-05,
"loss": 2.9067,
"step": 52500
},
{
"epoch": 0.88,
"learning_rate": 1.4157558520870193e-05,
"loss": 2.9148,
"step": 53000
},
{
"epoch": 0.88,
"learning_rate": 1.4102441148425574e-05,
"loss": 2.9172,
"step": 53500
},
{
"epoch": 0.89,
"learning_rate": 1.4047323775980952e-05,
"loss": 2.9063,
"step": 54000
},
{
"epoch": 0.9,
"learning_rate": 1.3992206403536333e-05,
"loss": 2.8933,
"step": 54500
},
{
"epoch": 0.91,
"learning_rate": 1.393708903109171e-05,
"loss": 2.9225,
"step": 55000
},
{
"epoch": 0.92,
"learning_rate": 1.3881971658647091e-05,
"loss": 2.9072,
"step": 55500
},
{
"epoch": 0.93,
"learning_rate": 1.3826854286202469e-05,
"loss": 2.909,
"step": 56000
},
{
"epoch": 0.93,
"learning_rate": 1.377173691375785e-05,
"loss": 2.8967,
"step": 56500
},
{
"epoch": 0.94,
"learning_rate": 1.3716619541313227e-05,
"loss": 2.915,
"step": 57000
},
{
"epoch": 0.95,
"learning_rate": 1.3661502168868608e-05,
"loss": 2.9118,
"step": 57500
},
{
"epoch": 0.96,
"learning_rate": 1.3606384796423986e-05,
"loss": 2.9086,
"step": 58000
},
{
"epoch": 0.97,
"learning_rate": 1.3551267423979367e-05,
"loss": 2.9099,
"step": 58500
},
{
"epoch": 0.98,
"learning_rate": 1.3496150051534744e-05,
"loss": 2.9073,
"step": 59000
},
{
"epoch": 0.98,
"learning_rate": 1.3441032679090122e-05,
"loss": 2.9182,
"step": 59500
},
{
"epoch": 0.99,
"learning_rate": 1.3385915306645503e-05,
"loss": 2.8954,
"step": 60000
}
],
"max_steps": 181431,
"num_train_epochs": 3,
"total_flos": 30196252016640000,
"trial_name": null,
"trial_params": null
}