m2m100_418M_ewe_fr_rel_news / trainer_state.json
Davlan's picture
add MT model
cbaeff3
raw
history blame
No virus
46.2 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"global_step": 186177,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 4.986571918120928e-05,
"loss": 2.7866,
"step": 500
},
{
"epoch": 0.02,
"learning_rate": 4.973143836241856e-05,
"loss": 2.4498,
"step": 1000
},
{
"epoch": 0.02,
"learning_rate": 4.959715754362784e-05,
"loss": 2.3135,
"step": 1500
},
{
"epoch": 0.03,
"learning_rate": 4.946287672483712e-05,
"loss": 2.2636,
"step": 2000
},
{
"epoch": 0.04,
"learning_rate": 4.93285959060464e-05,
"loss": 2.1826,
"step": 2500
},
{
"epoch": 0.05,
"learning_rate": 4.9194315087255675e-05,
"loss": 2.1323,
"step": 3000
},
{
"epoch": 0.06,
"learning_rate": 4.906003426846495e-05,
"loss": 2.0797,
"step": 3500
},
{
"epoch": 0.06,
"learning_rate": 4.8925753449674237e-05,
"loss": 2.0658,
"step": 4000
},
{
"epoch": 0.07,
"learning_rate": 4.8791472630883514e-05,
"loss": 2.0481,
"step": 4500
},
{
"epoch": 0.08,
"learning_rate": 4.865719181209279e-05,
"loss": 2.0033,
"step": 5000
},
{
"epoch": 0.09,
"learning_rate": 4.8522910993302076e-05,
"loss": 1.9998,
"step": 5500
},
{
"epoch": 0.1,
"learning_rate": 4.8388630174511354e-05,
"loss": 1.969,
"step": 6000
},
{
"epoch": 0.1,
"learning_rate": 4.825434935572064e-05,
"loss": 1.9782,
"step": 6500
},
{
"epoch": 0.11,
"learning_rate": 4.8120068536929916e-05,
"loss": 1.9142,
"step": 7000
},
{
"epoch": 0.12,
"learning_rate": 4.798578771813919e-05,
"loss": 1.9056,
"step": 7500
},
{
"epoch": 0.13,
"learning_rate": 4.785150689934847e-05,
"loss": 1.8963,
"step": 8000
},
{
"epoch": 0.14,
"learning_rate": 4.7717226080557755e-05,
"loss": 1.8883,
"step": 8500
},
{
"epoch": 0.15,
"learning_rate": 4.758294526176703e-05,
"loss": 1.8951,
"step": 9000
},
{
"epoch": 0.15,
"learning_rate": 4.744866444297631e-05,
"loss": 1.8463,
"step": 9500
},
{
"epoch": 0.16,
"learning_rate": 4.731438362418559e-05,
"loss": 1.8445,
"step": 10000
},
{
"epoch": 0.17,
"learning_rate": 4.7180102805394865e-05,
"loss": 1.8367,
"step": 10500
},
{
"epoch": 0.18,
"learning_rate": 4.704582198660415e-05,
"loss": 1.8117,
"step": 11000
},
{
"epoch": 0.19,
"learning_rate": 4.691154116781343e-05,
"loss": 1.8269,
"step": 11500
},
{
"epoch": 0.19,
"learning_rate": 4.6777260349022705e-05,
"loss": 1.8044,
"step": 12000
},
{
"epoch": 0.2,
"learning_rate": 4.664297953023198e-05,
"loss": 1.7914,
"step": 12500
},
{
"epoch": 0.21,
"learning_rate": 4.650869871144127e-05,
"loss": 1.7863,
"step": 13000
},
{
"epoch": 0.22,
"learning_rate": 4.6374417892650544e-05,
"loss": 1.7833,
"step": 13500
},
{
"epoch": 0.23,
"learning_rate": 4.624013707385982e-05,
"loss": 1.7548,
"step": 14000
},
{
"epoch": 0.23,
"learning_rate": 4.61058562550691e-05,
"loss": 1.7829,
"step": 14500
},
{
"epoch": 0.24,
"learning_rate": 4.597157543627838e-05,
"loss": 1.7706,
"step": 15000
},
{
"epoch": 0.25,
"learning_rate": 4.583729461748766e-05,
"loss": 1.7547,
"step": 15500
},
{
"epoch": 0.26,
"learning_rate": 4.570301379869694e-05,
"loss": 1.7549,
"step": 16000
},
{
"epoch": 0.27,
"learning_rate": 4.5568732979906224e-05,
"loss": 1.7475,
"step": 16500
},
{
"epoch": 0.27,
"learning_rate": 4.54344521611155e-05,
"loss": 1.7342,
"step": 17000
},
{
"epoch": 0.28,
"learning_rate": 4.530017134232478e-05,
"loss": 1.7228,
"step": 17500
},
{
"epoch": 0.29,
"learning_rate": 4.516589052353406e-05,
"loss": 1.7295,
"step": 18000
},
{
"epoch": 0.3,
"learning_rate": 4.503160970474334e-05,
"loss": 1.7197,
"step": 18500
},
{
"epoch": 0.31,
"learning_rate": 4.489732888595262e-05,
"loss": 1.7103,
"step": 19000
},
{
"epoch": 0.31,
"learning_rate": 4.4763048067161896e-05,
"loss": 1.7122,
"step": 19500
},
{
"epoch": 0.32,
"learning_rate": 4.462876724837118e-05,
"loss": 1.6975,
"step": 20000
},
{
"epoch": 0.33,
"learning_rate": 4.449448642958046e-05,
"loss": 1.715,
"step": 20500
},
{
"epoch": 0.34,
"learning_rate": 4.4360205610789735e-05,
"loss": 1.6843,
"step": 21000
},
{
"epoch": 0.35,
"learning_rate": 4.422592479199901e-05,
"loss": 1.6905,
"step": 21500
},
{
"epoch": 0.35,
"learning_rate": 4.409164397320829e-05,
"loss": 1.6821,
"step": 22000
},
{
"epoch": 0.36,
"learning_rate": 4.3957363154417575e-05,
"loss": 1.6892,
"step": 22500
},
{
"epoch": 0.37,
"learning_rate": 4.382308233562685e-05,
"loss": 1.6703,
"step": 23000
},
{
"epoch": 0.38,
"learning_rate": 4.368880151683613e-05,
"loss": 1.6664,
"step": 23500
},
{
"epoch": 0.39,
"learning_rate": 4.355452069804541e-05,
"loss": 1.6594,
"step": 24000
},
{
"epoch": 0.39,
"learning_rate": 4.3420239879254685e-05,
"loss": 1.6603,
"step": 24500
},
{
"epoch": 0.4,
"learning_rate": 4.328595906046397e-05,
"loss": 1.6624,
"step": 25000
},
{
"epoch": 0.41,
"learning_rate": 4.315167824167325e-05,
"loss": 1.6729,
"step": 25500
},
{
"epoch": 0.42,
"learning_rate": 4.3017397422882525e-05,
"loss": 1.6557,
"step": 26000
},
{
"epoch": 0.43,
"learning_rate": 4.28831166040918e-05,
"loss": 1.6494,
"step": 26500
},
{
"epoch": 0.44,
"learning_rate": 4.2748835785301087e-05,
"loss": 1.6262,
"step": 27000
},
{
"epoch": 0.44,
"learning_rate": 4.2614554966510364e-05,
"loss": 1.6452,
"step": 27500
},
{
"epoch": 0.45,
"learning_rate": 4.248027414771965e-05,
"loss": 1.6318,
"step": 28000
},
{
"epoch": 0.46,
"learning_rate": 4.2345993328928926e-05,
"loss": 1.6288,
"step": 28500
},
{
"epoch": 0.47,
"learning_rate": 4.2211712510138204e-05,
"loss": 1.6372,
"step": 29000
},
{
"epoch": 0.48,
"learning_rate": 4.207743169134749e-05,
"loss": 1.6327,
"step": 29500
},
{
"epoch": 0.48,
"learning_rate": 4.1943150872556766e-05,
"loss": 1.6249,
"step": 30000
},
{
"epoch": 0.49,
"learning_rate": 4.180887005376604e-05,
"loss": 1.6235,
"step": 30500
},
{
"epoch": 0.5,
"learning_rate": 4.167458923497532e-05,
"loss": 1.6402,
"step": 31000
},
{
"epoch": 0.51,
"learning_rate": 4.15403084161846e-05,
"loss": 1.6133,
"step": 31500
},
{
"epoch": 0.52,
"learning_rate": 4.140602759739388e-05,
"loss": 1.6075,
"step": 32000
},
{
"epoch": 0.52,
"learning_rate": 4.127174677860316e-05,
"loss": 1.6155,
"step": 32500
},
{
"epoch": 0.53,
"learning_rate": 4.113746595981244e-05,
"loss": 1.5978,
"step": 33000
},
{
"epoch": 0.54,
"learning_rate": 4.1003185141021715e-05,
"loss": 1.6027,
"step": 33500
},
{
"epoch": 0.55,
"learning_rate": 4.0868904322231e-05,
"loss": 1.6226,
"step": 34000
},
{
"epoch": 0.56,
"learning_rate": 4.073462350344028e-05,
"loss": 1.5909,
"step": 34500
},
{
"epoch": 0.56,
"learning_rate": 4.0600342684649555e-05,
"loss": 1.6034,
"step": 35000
},
{
"epoch": 0.57,
"learning_rate": 4.046606186585883e-05,
"loss": 1.578,
"step": 35500
},
{
"epoch": 0.58,
"learning_rate": 4.033178104706811e-05,
"loss": 1.5846,
"step": 36000
},
{
"epoch": 0.59,
"learning_rate": 4.0197500228277394e-05,
"loss": 1.5811,
"step": 36500
},
{
"epoch": 0.6,
"learning_rate": 4.006321940948667e-05,
"loss": 1.5719,
"step": 37000
},
{
"epoch": 0.6,
"learning_rate": 3.992893859069595e-05,
"loss": 1.5697,
"step": 37500
},
{
"epoch": 0.61,
"learning_rate": 3.979465777190523e-05,
"loss": 1.5685,
"step": 38000
},
{
"epoch": 0.62,
"learning_rate": 3.966037695311451e-05,
"loss": 1.5585,
"step": 38500
},
{
"epoch": 0.63,
"learning_rate": 3.952609613432379e-05,
"loss": 1.5852,
"step": 39000
},
{
"epoch": 0.64,
"learning_rate": 3.9391815315533074e-05,
"loss": 1.5664,
"step": 39500
},
{
"epoch": 0.64,
"learning_rate": 3.925753449674235e-05,
"loss": 1.564,
"step": 40000
},
{
"epoch": 0.65,
"learning_rate": 3.912325367795163e-05,
"loss": 1.549,
"step": 40500
},
{
"epoch": 0.66,
"learning_rate": 3.898897285916091e-05,
"loss": 1.5394,
"step": 41000
},
{
"epoch": 0.67,
"learning_rate": 3.885469204037019e-05,
"loss": 1.5442,
"step": 41500
},
{
"epoch": 0.68,
"learning_rate": 3.872041122157947e-05,
"loss": 1.5577,
"step": 42000
},
{
"epoch": 0.68,
"learning_rate": 3.8586130402788746e-05,
"loss": 1.5327,
"step": 42500
},
{
"epoch": 0.69,
"learning_rate": 3.845184958399802e-05,
"loss": 1.5609,
"step": 43000
},
{
"epoch": 0.7,
"learning_rate": 3.831756876520731e-05,
"loss": 1.5704,
"step": 43500
},
{
"epoch": 0.71,
"learning_rate": 3.8183287946416585e-05,
"loss": 1.5513,
"step": 44000
},
{
"epoch": 0.72,
"learning_rate": 3.804900712762586e-05,
"loss": 1.5461,
"step": 44500
},
{
"epoch": 0.73,
"learning_rate": 3.791472630883514e-05,
"loss": 1.5198,
"step": 45000
},
{
"epoch": 0.73,
"learning_rate": 3.778044549004442e-05,
"loss": 1.5184,
"step": 45500
},
{
"epoch": 0.74,
"learning_rate": 3.76461646712537e-05,
"loss": 1.5497,
"step": 46000
},
{
"epoch": 0.75,
"learning_rate": 3.751188385246298e-05,
"loss": 1.5488,
"step": 46500
},
{
"epoch": 0.76,
"learning_rate": 3.737760303367226e-05,
"loss": 1.5189,
"step": 47000
},
{
"epoch": 0.77,
"learning_rate": 3.7243322214881535e-05,
"loss": 1.5302,
"step": 47500
},
{
"epoch": 0.77,
"learning_rate": 3.710904139609082e-05,
"loss": 1.5082,
"step": 48000
},
{
"epoch": 0.78,
"learning_rate": 3.69747605773001e-05,
"loss": 1.53,
"step": 48500
},
{
"epoch": 0.79,
"learning_rate": 3.6840479758509375e-05,
"loss": 1.511,
"step": 49000
},
{
"epoch": 0.8,
"learning_rate": 3.670619893971866e-05,
"loss": 1.5137,
"step": 49500
},
{
"epoch": 0.81,
"learning_rate": 3.6571918120927937e-05,
"loss": 1.526,
"step": 50000
},
{
"epoch": 0.81,
"learning_rate": 3.6437637302137214e-05,
"loss": 1.5175,
"step": 50500
},
{
"epoch": 0.82,
"learning_rate": 3.63033564833465e-05,
"loss": 1.4989,
"step": 51000
},
{
"epoch": 0.83,
"learning_rate": 3.6169075664555776e-05,
"loss": 1.5158,
"step": 51500
},
{
"epoch": 0.84,
"learning_rate": 3.6034794845765054e-05,
"loss": 1.4929,
"step": 52000
},
{
"epoch": 0.85,
"learning_rate": 3.590051402697433e-05,
"loss": 1.4921,
"step": 52500
},
{
"epoch": 0.85,
"learning_rate": 3.5766233208183616e-05,
"loss": 1.5086,
"step": 53000
},
{
"epoch": 0.86,
"learning_rate": 3.563195238939289e-05,
"loss": 1.5026,
"step": 53500
},
{
"epoch": 0.87,
"learning_rate": 3.549767157060217e-05,
"loss": 1.4813,
"step": 54000
},
{
"epoch": 0.88,
"learning_rate": 3.536339075181145e-05,
"loss": 1.5041,
"step": 54500
},
{
"epoch": 0.89,
"learning_rate": 3.522910993302073e-05,
"loss": 1.5063,
"step": 55000
},
{
"epoch": 0.89,
"learning_rate": 3.509482911423001e-05,
"loss": 1.49,
"step": 55500
},
{
"epoch": 0.9,
"learning_rate": 3.496054829543929e-05,
"loss": 1.4972,
"step": 56000
},
{
"epoch": 0.91,
"learning_rate": 3.4826267476648565e-05,
"loss": 1.4701,
"step": 56500
},
{
"epoch": 0.92,
"learning_rate": 3.469198665785784e-05,
"loss": 1.4791,
"step": 57000
},
{
"epoch": 0.93,
"learning_rate": 3.455770583906713e-05,
"loss": 1.484,
"step": 57500
},
{
"epoch": 0.93,
"learning_rate": 3.4423425020276405e-05,
"loss": 1.4761,
"step": 58000
},
{
"epoch": 0.94,
"learning_rate": 3.428914420148568e-05,
"loss": 1.4697,
"step": 58500
},
{
"epoch": 0.95,
"learning_rate": 3.415486338269496e-05,
"loss": 1.475,
"step": 59000
},
{
"epoch": 0.96,
"learning_rate": 3.402058256390424e-05,
"loss": 1.4788,
"step": 59500
},
{
"epoch": 0.97,
"learning_rate": 3.388630174511352e-05,
"loss": 1.468,
"step": 60000
},
{
"epoch": 0.97,
"learning_rate": 3.37520209263228e-05,
"loss": 1.4903,
"step": 60500
},
{
"epoch": 0.98,
"learning_rate": 3.3617740107532084e-05,
"loss": 1.442,
"step": 61000
},
{
"epoch": 0.99,
"learning_rate": 3.348345928874136e-05,
"loss": 1.4327,
"step": 61500
},
{
"epoch": 1.0,
"learning_rate": 3.3349178469950646e-05,
"loss": 1.4776,
"step": 62000
},
{
"epoch": 1.01,
"learning_rate": 3.3214897651159924e-05,
"loss": 1.3768,
"step": 62500
},
{
"epoch": 1.02,
"learning_rate": 3.30806168323692e-05,
"loss": 1.3501,
"step": 63000
},
{
"epoch": 1.02,
"learning_rate": 3.294633601357848e-05,
"loss": 1.3455,
"step": 63500
},
{
"epoch": 1.03,
"learning_rate": 3.2812055194787756e-05,
"loss": 1.3517,
"step": 64000
},
{
"epoch": 1.04,
"learning_rate": 3.267777437599704e-05,
"loss": 1.3592,
"step": 64500
},
{
"epoch": 1.05,
"learning_rate": 3.254349355720632e-05,
"loss": 1.3677,
"step": 65000
},
{
"epoch": 1.06,
"learning_rate": 3.2409212738415596e-05,
"loss": 1.372,
"step": 65500
},
{
"epoch": 1.06,
"learning_rate": 3.227493191962487e-05,
"loss": 1.349,
"step": 66000
},
{
"epoch": 1.07,
"learning_rate": 3.214065110083415e-05,
"loss": 1.3581,
"step": 66500
},
{
"epoch": 1.08,
"learning_rate": 3.2006370282043435e-05,
"loss": 1.3444,
"step": 67000
},
{
"epoch": 1.09,
"learning_rate": 3.187208946325271e-05,
"loss": 1.3505,
"step": 67500
},
{
"epoch": 1.1,
"learning_rate": 3.173780864446199e-05,
"loss": 1.347,
"step": 68000
},
{
"epoch": 1.1,
"learning_rate": 3.160352782567127e-05,
"loss": 1.3501,
"step": 68500
},
{
"epoch": 1.11,
"learning_rate": 3.146924700688055e-05,
"loss": 1.3432,
"step": 69000
},
{
"epoch": 1.12,
"learning_rate": 3.133496618808983e-05,
"loss": 1.3576,
"step": 69500
},
{
"epoch": 1.13,
"learning_rate": 3.120068536929911e-05,
"loss": 1.3497,
"step": 70000
},
{
"epoch": 1.14,
"learning_rate": 3.1066404550508385e-05,
"loss": 1.3417,
"step": 70500
},
{
"epoch": 1.14,
"learning_rate": 3.093212373171766e-05,
"loss": 1.3521,
"step": 71000
},
{
"epoch": 1.15,
"learning_rate": 3.079784291292695e-05,
"loss": 1.3484,
"step": 71500
},
{
"epoch": 1.16,
"learning_rate": 3.0663562094136225e-05,
"loss": 1.3295,
"step": 72000
},
{
"epoch": 1.17,
"learning_rate": 3.052928127534551e-05,
"loss": 1.3487,
"step": 72500
},
{
"epoch": 1.18,
"learning_rate": 3.0395000456554783e-05,
"loss": 1.3581,
"step": 73000
},
{
"epoch": 1.18,
"learning_rate": 3.026071963776406e-05,
"loss": 1.3505,
"step": 73500
},
{
"epoch": 1.19,
"learning_rate": 3.0126438818973345e-05,
"loss": 1.3413,
"step": 74000
},
{
"epoch": 1.2,
"learning_rate": 2.9992158000182623e-05,
"loss": 1.3559,
"step": 74500
},
{
"epoch": 1.21,
"learning_rate": 2.9857877181391904e-05,
"loss": 1.3324,
"step": 75000
},
{
"epoch": 1.22,
"learning_rate": 2.972359636260118e-05,
"loss": 1.3446,
"step": 75500
},
{
"epoch": 1.22,
"learning_rate": 2.9589315543810466e-05,
"loss": 1.3608,
"step": 76000
},
{
"epoch": 1.23,
"learning_rate": 2.9455034725019743e-05,
"loss": 1.3405,
"step": 76500
},
{
"epoch": 1.24,
"learning_rate": 2.932075390622902e-05,
"loss": 1.3385,
"step": 77000
},
{
"epoch": 1.25,
"learning_rate": 2.91864730874383e-05,
"loss": 1.3236,
"step": 77500
},
{
"epoch": 1.26,
"learning_rate": 2.9052192268647576e-05,
"loss": 1.3478,
"step": 78000
},
{
"epoch": 1.26,
"learning_rate": 2.891791144985686e-05,
"loss": 1.3259,
"step": 78500
},
{
"epoch": 1.27,
"learning_rate": 2.8783630631066138e-05,
"loss": 1.3126,
"step": 79000
},
{
"epoch": 1.28,
"learning_rate": 2.8649349812275415e-05,
"loss": 1.3269,
"step": 79500
},
{
"epoch": 1.29,
"learning_rate": 2.8515068993484696e-05,
"loss": 1.356,
"step": 80000
},
{
"epoch": 1.3,
"learning_rate": 2.8380788174693974e-05,
"loss": 1.3278,
"step": 80500
},
{
"epoch": 1.31,
"learning_rate": 2.824650735590326e-05,
"loss": 1.327,
"step": 81000
},
{
"epoch": 1.31,
"learning_rate": 2.8112226537112536e-05,
"loss": 1.3308,
"step": 81500
},
{
"epoch": 1.32,
"learning_rate": 2.7977945718321814e-05,
"loss": 1.3526,
"step": 82000
},
{
"epoch": 1.33,
"learning_rate": 2.784366489953109e-05,
"loss": 1.3455,
"step": 82500
},
{
"epoch": 1.34,
"learning_rate": 2.770938408074037e-05,
"loss": 1.3159,
"step": 83000
},
{
"epoch": 1.35,
"learning_rate": 2.7575103261949653e-05,
"loss": 1.328,
"step": 83500
},
{
"epoch": 1.35,
"learning_rate": 2.744082244315893e-05,
"loss": 1.3177,
"step": 84000
},
{
"epoch": 1.36,
"learning_rate": 2.7306541624368208e-05,
"loss": 1.3269,
"step": 84500
},
{
"epoch": 1.37,
"learning_rate": 2.7172260805577486e-05,
"loss": 1.3302,
"step": 85000
},
{
"epoch": 1.38,
"learning_rate": 2.703797998678677e-05,
"loss": 1.3337,
"step": 85500
},
{
"epoch": 1.39,
"learning_rate": 2.6903699167996048e-05,
"loss": 1.3212,
"step": 86000
},
{
"epoch": 1.39,
"learning_rate": 2.676941834920533e-05,
"loss": 1.3213,
"step": 86500
},
{
"epoch": 1.4,
"learning_rate": 2.6635137530414606e-05,
"loss": 1.3306,
"step": 87000
},
{
"epoch": 1.41,
"learning_rate": 2.6500856711623884e-05,
"loss": 1.3412,
"step": 87500
},
{
"epoch": 1.42,
"learning_rate": 2.6366575892833168e-05,
"loss": 1.3054,
"step": 88000
},
{
"epoch": 1.43,
"learning_rate": 2.6232295074042446e-05,
"loss": 1.303,
"step": 88500
},
{
"epoch": 1.43,
"learning_rate": 2.6098014255251723e-05,
"loss": 1.3191,
"step": 89000
},
{
"epoch": 1.44,
"learning_rate": 2.5963733436461e-05,
"loss": 1.3208,
"step": 89500
},
{
"epoch": 1.45,
"learning_rate": 2.582945261767028e-05,
"loss": 1.3288,
"step": 90000
},
{
"epoch": 1.46,
"learning_rate": 2.5695171798879563e-05,
"loss": 1.3477,
"step": 90500
},
{
"epoch": 1.47,
"learning_rate": 2.556089098008884e-05,
"loss": 1.3199,
"step": 91000
},
{
"epoch": 1.47,
"learning_rate": 2.542661016129812e-05,
"loss": 1.3148,
"step": 91500
},
{
"epoch": 1.48,
"learning_rate": 2.52923293425074e-05,
"loss": 1.3203,
"step": 92000
},
{
"epoch": 1.49,
"learning_rate": 2.5158048523716683e-05,
"loss": 1.3177,
"step": 92500
},
{
"epoch": 1.5,
"learning_rate": 2.502376770492596e-05,
"loss": 1.3275,
"step": 93000
},
{
"epoch": 1.51,
"learning_rate": 2.488948688613524e-05,
"loss": 1.3226,
"step": 93500
},
{
"epoch": 1.51,
"learning_rate": 2.4755206067344516e-05,
"loss": 1.3091,
"step": 94000
},
{
"epoch": 1.52,
"learning_rate": 2.4620925248553797e-05,
"loss": 1.3163,
"step": 94500
},
{
"epoch": 1.53,
"learning_rate": 2.4486644429763075e-05,
"loss": 1.3065,
"step": 95000
},
{
"epoch": 1.54,
"learning_rate": 2.4352363610972356e-05,
"loss": 1.3021,
"step": 95500
},
{
"epoch": 1.55,
"learning_rate": 2.4218082792181633e-05,
"loss": 1.3256,
"step": 96000
},
{
"epoch": 1.55,
"learning_rate": 2.4083801973390914e-05,
"loss": 1.3157,
"step": 96500
},
{
"epoch": 1.56,
"learning_rate": 2.3949521154600195e-05,
"loss": 1.3202,
"step": 97000
},
{
"epoch": 1.57,
"learning_rate": 2.3815240335809473e-05,
"loss": 1.3099,
"step": 97500
},
{
"epoch": 1.58,
"learning_rate": 2.3680959517018754e-05,
"loss": 1.3008,
"step": 98000
},
{
"epoch": 1.59,
"learning_rate": 2.354667869822803e-05,
"loss": 1.312,
"step": 98500
},
{
"epoch": 1.6,
"learning_rate": 2.3412397879437312e-05,
"loss": 1.3168,
"step": 99000
},
{
"epoch": 1.6,
"learning_rate": 2.327811706064659e-05,
"loss": 1.312,
"step": 99500
},
{
"epoch": 1.61,
"learning_rate": 2.314383624185587e-05,
"loss": 1.2771,
"step": 100000
},
{
"epoch": 1.62,
"learning_rate": 2.300955542306515e-05,
"loss": 1.2747,
"step": 100500
},
{
"epoch": 1.63,
"learning_rate": 2.2875274604274426e-05,
"loss": 1.3005,
"step": 101000
},
{
"epoch": 1.64,
"learning_rate": 2.2740993785483707e-05,
"loss": 1.2913,
"step": 101500
},
{
"epoch": 1.64,
"learning_rate": 2.2606712966692984e-05,
"loss": 1.2809,
"step": 102000
},
{
"epoch": 1.65,
"learning_rate": 2.2472432147902265e-05,
"loss": 1.3009,
"step": 102500
},
{
"epoch": 1.66,
"learning_rate": 2.2338151329111546e-05,
"loss": 1.2823,
"step": 103000
},
{
"epoch": 1.67,
"learning_rate": 2.2203870510320827e-05,
"loss": 1.2675,
"step": 103500
},
{
"epoch": 1.68,
"learning_rate": 2.2069589691530105e-05,
"loss": 1.2909,
"step": 104000
},
{
"epoch": 1.68,
"learning_rate": 2.1935308872739383e-05,
"loss": 1.2863,
"step": 104500
},
{
"epoch": 1.69,
"learning_rate": 2.1801028053948664e-05,
"loss": 1.286,
"step": 105000
},
{
"epoch": 1.7,
"learning_rate": 2.166674723515794e-05,
"loss": 1.282,
"step": 105500
},
{
"epoch": 1.71,
"learning_rate": 2.1532466416367222e-05,
"loss": 1.296,
"step": 106000
},
{
"epoch": 1.72,
"learning_rate": 2.13981855975765e-05,
"loss": 1.2833,
"step": 106500
},
{
"epoch": 1.72,
"learning_rate": 2.126390477878578e-05,
"loss": 1.2802,
"step": 107000
},
{
"epoch": 1.73,
"learning_rate": 2.1129623959995058e-05,
"loss": 1.2949,
"step": 107500
},
{
"epoch": 1.74,
"learning_rate": 2.099534314120434e-05,
"loss": 1.2805,
"step": 108000
},
{
"epoch": 1.75,
"learning_rate": 2.086106232241362e-05,
"loss": 1.2932,
"step": 108500
},
{
"epoch": 1.76,
"learning_rate": 2.0726781503622898e-05,
"loss": 1.288,
"step": 109000
},
{
"epoch": 1.76,
"learning_rate": 2.059250068483218e-05,
"loss": 1.2868,
"step": 109500
},
{
"epoch": 1.77,
"learning_rate": 2.0458219866041456e-05,
"loss": 1.284,
"step": 110000
},
{
"epoch": 1.78,
"learning_rate": 2.0323939047250737e-05,
"loss": 1.2842,
"step": 110500
},
{
"epoch": 1.79,
"learning_rate": 2.0189658228460015e-05,
"loss": 1.2893,
"step": 111000
},
{
"epoch": 1.8,
"learning_rate": 2.0055377409669292e-05,
"loss": 1.273,
"step": 111500
},
{
"epoch": 1.8,
"learning_rate": 1.9921096590878573e-05,
"loss": 1.2849,
"step": 112000
},
{
"epoch": 1.81,
"learning_rate": 1.978681577208785e-05,
"loss": 1.2844,
"step": 112500
},
{
"epoch": 1.82,
"learning_rate": 1.9652534953297132e-05,
"loss": 1.2726,
"step": 113000
},
{
"epoch": 1.83,
"learning_rate": 1.9518254134506413e-05,
"loss": 1.2635,
"step": 113500
},
{
"epoch": 1.84,
"learning_rate": 1.9383973315715694e-05,
"loss": 1.2784,
"step": 114000
},
{
"epoch": 1.85,
"learning_rate": 1.924969249692497e-05,
"loss": 1.2515,
"step": 114500
},
{
"epoch": 1.85,
"learning_rate": 1.911541167813425e-05,
"loss": 1.2878,
"step": 115000
},
{
"epoch": 1.86,
"learning_rate": 1.898113085934353e-05,
"loss": 1.2453,
"step": 115500
},
{
"epoch": 1.87,
"learning_rate": 1.8846850040552808e-05,
"loss": 1.2749,
"step": 116000
},
{
"epoch": 1.88,
"learning_rate": 1.871256922176209e-05,
"loss": 1.2798,
"step": 116500
},
{
"epoch": 1.89,
"learning_rate": 1.8578288402971366e-05,
"loss": 1.2728,
"step": 117000
},
{
"epoch": 1.89,
"learning_rate": 1.8444007584180647e-05,
"loss": 1.2657,
"step": 117500
},
{
"epoch": 1.9,
"learning_rate": 1.8309726765389925e-05,
"loss": 1.284,
"step": 118000
},
{
"epoch": 1.91,
"learning_rate": 1.8175445946599202e-05,
"loss": 1.2842,
"step": 118500
},
{
"epoch": 1.92,
"learning_rate": 1.8041165127808483e-05,
"loss": 1.2627,
"step": 119000
},
{
"epoch": 1.93,
"learning_rate": 1.7906884309017764e-05,
"loss": 1.2529,
"step": 119500
},
{
"epoch": 1.93,
"learning_rate": 1.7772603490227045e-05,
"loss": 1.2632,
"step": 120000
},
{
"epoch": 1.94,
"learning_rate": 1.7638322671436323e-05,
"loss": 1.2493,
"step": 120500
},
{
"epoch": 1.95,
"learning_rate": 1.75040418526456e-05,
"loss": 1.2633,
"step": 121000
},
{
"epoch": 1.96,
"learning_rate": 1.736976103385488e-05,
"loss": 1.2698,
"step": 121500
},
{
"epoch": 1.97,
"learning_rate": 1.723548021506416e-05,
"loss": 1.2574,
"step": 122000
},
{
"epoch": 1.97,
"learning_rate": 1.710119939627344e-05,
"loss": 1.2613,
"step": 122500
},
{
"epoch": 1.98,
"learning_rate": 1.6966918577482717e-05,
"loss": 1.2436,
"step": 123000
},
{
"epoch": 1.99,
"learning_rate": 1.6832637758692e-05,
"loss": 1.2495,
"step": 123500
},
{
"epoch": 2.0,
"learning_rate": 1.6698356939901276e-05,
"loss": 1.2633,
"step": 124000
},
{
"epoch": 2.01,
"learning_rate": 1.6564076121110557e-05,
"loss": 1.1792,
"step": 124500
},
{
"epoch": 2.01,
"learning_rate": 1.6429795302319838e-05,
"loss": 1.1426,
"step": 125000
},
{
"epoch": 2.02,
"learning_rate": 1.6295514483529115e-05,
"loss": 1.141,
"step": 125500
},
{
"epoch": 2.03,
"learning_rate": 1.6161233664738396e-05,
"loss": 1.1474,
"step": 126000
},
{
"epoch": 2.04,
"learning_rate": 1.6026952845947674e-05,
"loss": 1.1307,
"step": 126500
},
{
"epoch": 2.05,
"learning_rate": 1.5892672027156955e-05,
"loss": 1.1471,
"step": 127000
},
{
"epoch": 2.05,
"learning_rate": 1.5758391208366233e-05,
"loss": 1.1523,
"step": 127500
},
{
"epoch": 2.06,
"learning_rate": 1.562411038957551e-05,
"loss": 1.1379,
"step": 128000
},
{
"epoch": 2.07,
"learning_rate": 1.548982957078479e-05,
"loss": 1.1391,
"step": 128500
},
{
"epoch": 2.08,
"learning_rate": 1.535554875199407e-05,
"loss": 1.1483,
"step": 129000
},
{
"epoch": 2.09,
"learning_rate": 1.5221267933203351e-05,
"loss": 1.1214,
"step": 129500
},
{
"epoch": 2.09,
"learning_rate": 1.5086987114412629e-05,
"loss": 1.1369,
"step": 130000
},
{
"epoch": 2.1,
"learning_rate": 1.495270629562191e-05,
"loss": 1.1431,
"step": 130500
},
{
"epoch": 2.11,
"learning_rate": 1.4818425476831187e-05,
"loss": 1.1372,
"step": 131000
},
{
"epoch": 2.12,
"learning_rate": 1.4684144658040467e-05,
"loss": 1.1464,
"step": 131500
},
{
"epoch": 2.13,
"learning_rate": 1.4549863839249748e-05,
"loss": 1.1253,
"step": 132000
},
{
"epoch": 2.14,
"learning_rate": 1.4415583020459025e-05,
"loss": 1.1317,
"step": 132500
},
{
"epoch": 2.14,
"learning_rate": 1.4281302201668306e-05,
"loss": 1.1529,
"step": 133000
},
{
"epoch": 2.15,
"learning_rate": 1.4147021382877584e-05,
"loss": 1.165,
"step": 133500
},
{
"epoch": 2.16,
"learning_rate": 1.4012740564086865e-05,
"loss": 1.1055,
"step": 134000
},
{
"epoch": 2.17,
"learning_rate": 1.3878459745296144e-05,
"loss": 1.1556,
"step": 134500
},
{
"epoch": 2.18,
"learning_rate": 1.3744178926505422e-05,
"loss": 1.1296,
"step": 135000
},
{
"epoch": 2.18,
"learning_rate": 1.3609898107714703e-05,
"loss": 1.1462,
"step": 135500
},
{
"epoch": 2.19,
"learning_rate": 1.347561728892398e-05,
"loss": 1.134,
"step": 136000
},
{
"epoch": 2.2,
"learning_rate": 1.3341336470133261e-05,
"loss": 1.1417,
"step": 136500
},
{
"epoch": 2.21,
"learning_rate": 1.320705565134254e-05,
"loss": 1.145,
"step": 137000
},
{
"epoch": 2.22,
"learning_rate": 1.3072774832551821e-05,
"loss": 1.1229,
"step": 137500
},
{
"epoch": 2.22,
"learning_rate": 1.2938494013761099e-05,
"loss": 1.1339,
"step": 138000
},
{
"epoch": 2.23,
"learning_rate": 1.2804213194970377e-05,
"loss": 1.1418,
"step": 138500
},
{
"epoch": 2.24,
"learning_rate": 1.2669932376179658e-05,
"loss": 1.1331,
"step": 139000
},
{
"epoch": 2.25,
"learning_rate": 1.2535651557388937e-05,
"loss": 1.1242,
"step": 139500
},
{
"epoch": 2.26,
"learning_rate": 1.2401370738598218e-05,
"loss": 1.1322,
"step": 140000
},
{
"epoch": 2.26,
"learning_rate": 1.2267089919807495e-05,
"loss": 1.1367,
"step": 140500
},
{
"epoch": 2.27,
"learning_rate": 1.2132809101016775e-05,
"loss": 1.1419,
"step": 141000
},
{
"epoch": 2.28,
"learning_rate": 1.1998528282226054e-05,
"loss": 1.1381,
"step": 141500
},
{
"epoch": 2.29,
"learning_rate": 1.1864247463435333e-05,
"loss": 1.1529,
"step": 142000
},
{
"epoch": 2.3,
"learning_rate": 1.1729966644644612e-05,
"loss": 1.1315,
"step": 142500
},
{
"epoch": 2.3,
"learning_rate": 1.1595685825853893e-05,
"loss": 1.1399,
"step": 143000
},
{
"epoch": 2.31,
"learning_rate": 1.1461405007063171e-05,
"loss": 1.124,
"step": 143500
},
{
"epoch": 2.32,
"learning_rate": 1.132712418827245e-05,
"loss": 1.1339,
"step": 144000
},
{
"epoch": 2.33,
"learning_rate": 1.119284336948173e-05,
"loss": 1.1269,
"step": 144500
},
{
"epoch": 2.34,
"learning_rate": 1.1058562550691009e-05,
"loss": 1.1311,
"step": 145000
},
{
"epoch": 2.34,
"learning_rate": 1.092428173190029e-05,
"loss": 1.1458,
"step": 145500
},
{
"epoch": 2.35,
"learning_rate": 1.0790000913109569e-05,
"loss": 1.1142,
"step": 146000
},
{
"epoch": 2.36,
"learning_rate": 1.0655720094318848e-05,
"loss": 1.1144,
"step": 146500
},
{
"epoch": 2.37,
"learning_rate": 1.0521439275528126e-05,
"loss": 1.139,
"step": 147000
},
{
"epoch": 2.38,
"learning_rate": 1.0387158456737405e-05,
"loss": 1.1229,
"step": 147500
},
{
"epoch": 2.38,
"learning_rate": 1.0252877637946686e-05,
"loss": 1.1285,
"step": 148000
},
{
"epoch": 2.39,
"learning_rate": 1.0118596819155965e-05,
"loss": 1.1211,
"step": 148500
},
{
"epoch": 2.4,
"learning_rate": 9.984316000365245e-06,
"loss": 1.1423,
"step": 149000
},
{
"epoch": 2.41,
"learning_rate": 9.850035181574524e-06,
"loss": 1.117,
"step": 149500
},
{
"epoch": 2.42,
"learning_rate": 9.715754362783803e-06,
"loss": 1.1299,
"step": 150000
},
{
"epoch": 2.43,
"learning_rate": 9.58147354399308e-06,
"loss": 1.1213,
"step": 150500
},
{
"epoch": 2.43,
"learning_rate": 9.447192725202362e-06,
"loss": 1.1124,
"step": 151000
},
{
"epoch": 2.44,
"learning_rate": 9.312911906411641e-06,
"loss": 1.1147,
"step": 151500
},
{
"epoch": 2.45,
"learning_rate": 9.17863108762092e-06,
"loss": 1.1119,
"step": 152000
},
{
"epoch": 2.46,
"learning_rate": 9.0443502688302e-06,
"loss": 1.118,
"step": 152500
},
{
"epoch": 2.47,
"learning_rate": 8.910069450039479e-06,
"loss": 1.1321,
"step": 153000
},
{
"epoch": 2.47,
"learning_rate": 8.775788631248758e-06,
"loss": 1.1259,
"step": 153500
},
{
"epoch": 2.48,
"learning_rate": 8.641507812458037e-06,
"loss": 1.1178,
"step": 154000
},
{
"epoch": 2.49,
"learning_rate": 8.507226993667317e-06,
"loss": 1.1344,
"step": 154500
},
{
"epoch": 2.5,
"learning_rate": 8.372946174876596e-06,
"loss": 1.1059,
"step": 155000
},
{
"epoch": 2.51,
"learning_rate": 8.238665356085875e-06,
"loss": 1.123,
"step": 155500
},
{
"epoch": 2.51,
"learning_rate": 8.104384537295155e-06,
"loss": 1.128,
"step": 156000
},
{
"epoch": 2.52,
"learning_rate": 7.970103718504436e-06,
"loss": 1.124,
"step": 156500
},
{
"epoch": 2.53,
"learning_rate": 7.835822899713715e-06,
"loss": 1.1264,
"step": 157000
},
{
"epoch": 2.54,
"learning_rate": 7.701542080922992e-06,
"loss": 1.1084,
"step": 157500
},
{
"epoch": 2.55,
"learning_rate": 7.567261262132272e-06,
"loss": 1.1307,
"step": 158000
},
{
"epoch": 2.55,
"learning_rate": 7.432980443341552e-06,
"loss": 1.1095,
"step": 158500
},
{
"epoch": 2.56,
"learning_rate": 7.298699624550831e-06,
"loss": 1.1145,
"step": 159000
},
{
"epoch": 2.57,
"learning_rate": 7.16441880576011e-06,
"loss": 1.1163,
"step": 159500
},
{
"epoch": 2.58,
"learning_rate": 7.0301379869693904e-06,
"loss": 1.1303,
"step": 160000
},
{
"epoch": 2.59,
"learning_rate": 6.89585716817867e-06,
"loss": 1.1204,
"step": 160500
},
{
"epoch": 2.59,
"learning_rate": 6.761576349387948e-06,
"loss": 1.0957,
"step": 161000
},
{
"epoch": 2.6,
"learning_rate": 6.627295530597227e-06,
"loss": 1.1141,
"step": 161500
},
{
"epoch": 2.61,
"learning_rate": 6.493014711806507e-06,
"loss": 1.111,
"step": 162000
},
{
"epoch": 2.62,
"learning_rate": 6.358733893015786e-06,
"loss": 1.1059,
"step": 162500
},
{
"epoch": 2.63,
"learning_rate": 6.224453074225066e-06,
"loss": 1.1088,
"step": 163000
},
{
"epoch": 2.63,
"learning_rate": 6.0901722554343445e-06,
"loss": 1.127,
"step": 163500
},
{
"epoch": 2.64,
"learning_rate": 5.955891436643625e-06,
"loss": 1.1217,
"step": 164000
},
{
"epoch": 2.65,
"learning_rate": 5.821610617852904e-06,
"loss": 1.1208,
"step": 164500
},
{
"epoch": 2.66,
"learning_rate": 5.687329799062183e-06,
"loss": 1.1139,
"step": 165000
},
{
"epoch": 2.67,
"learning_rate": 5.5530489802714624e-06,
"loss": 1.1268,
"step": 165500
},
{
"epoch": 2.67,
"learning_rate": 5.418768161480742e-06,
"loss": 1.0985,
"step": 166000
},
{
"epoch": 2.68,
"learning_rate": 5.284487342690021e-06,
"loss": 1.11,
"step": 166500
},
{
"epoch": 2.69,
"learning_rate": 5.1502065238993e-06,
"loss": 1.1408,
"step": 167000
},
{
"epoch": 2.7,
"learning_rate": 5.0159257051085795e-06,
"loss": 1.1029,
"step": 167500
},
{
"epoch": 2.71,
"learning_rate": 4.881644886317859e-06,
"loss": 1.1006,
"step": 168000
},
{
"epoch": 2.72,
"learning_rate": 4.747364067527139e-06,
"loss": 1.1177,
"step": 168500
},
{
"epoch": 2.72,
"learning_rate": 4.613083248736417e-06,
"loss": 1.104,
"step": 169000
},
{
"epoch": 2.73,
"learning_rate": 4.4788024299456975e-06,
"loss": 1.0976,
"step": 169500
},
{
"epoch": 2.74,
"learning_rate": 4.344521611154977e-06,
"loss": 1.1053,
"step": 170000
},
{
"epoch": 2.75,
"learning_rate": 4.210240792364255e-06,
"loss": 1.1197,
"step": 170500
},
{
"epoch": 2.76,
"learning_rate": 4.075959973573535e-06,
"loss": 1.1158,
"step": 171000
},
{
"epoch": 2.76,
"learning_rate": 3.941679154782815e-06,
"loss": 1.1145,
"step": 171500
},
{
"epoch": 2.77,
"learning_rate": 3.8073983359920934e-06,
"loss": 1.095,
"step": 172000
},
{
"epoch": 2.78,
"learning_rate": 3.673117517201373e-06,
"loss": 1.1133,
"step": 172500
},
{
"epoch": 2.79,
"learning_rate": 3.5388366984106524e-06,
"loss": 1.0898,
"step": 173000
},
{
"epoch": 2.8,
"learning_rate": 3.404555879619932e-06,
"loss": 1.1011,
"step": 173500
},
{
"epoch": 2.8,
"learning_rate": 3.270275060829211e-06,
"loss": 1.0964,
"step": 174000
},
{
"epoch": 2.81,
"learning_rate": 3.1359942420384902e-06,
"loss": 1.1101,
"step": 174500
},
{
"epoch": 2.82,
"learning_rate": 3.00171342324777e-06,
"loss": 1.1091,
"step": 175000
},
{
"epoch": 2.83,
"learning_rate": 2.867432604457049e-06,
"loss": 1.0885,
"step": 175500
},
{
"epoch": 2.84,
"learning_rate": 2.7331517856663285e-06,
"loss": 1.0921,
"step": 176000
},
{
"epoch": 2.84,
"learning_rate": 2.5988709668756078e-06,
"loss": 1.1216,
"step": 176500
},
{
"epoch": 2.85,
"learning_rate": 2.464590148084887e-06,
"loss": 1.1013,
"step": 177000
},
{
"epoch": 2.86,
"learning_rate": 2.3303093292941667e-06,
"loss": 1.0844,
"step": 177500
},
{
"epoch": 2.87,
"learning_rate": 2.1960285105034456e-06,
"loss": 1.1088,
"step": 178000
},
{
"epoch": 2.88,
"learning_rate": 2.061747691712725e-06,
"loss": 1.113,
"step": 178500
},
{
"epoch": 2.88,
"learning_rate": 1.9274668729220045e-06,
"loss": 1.0921,
"step": 179000
},
{
"epoch": 2.89,
"learning_rate": 1.7931860541312836e-06,
"loss": 1.1002,
"step": 179500
},
{
"epoch": 2.9,
"learning_rate": 1.658905235340563e-06,
"loss": 1.1014,
"step": 180000
},
{
"epoch": 2.91,
"learning_rate": 1.5246244165498424e-06,
"loss": 1.1051,
"step": 180500
},
{
"epoch": 2.92,
"learning_rate": 1.3903435977591216e-06,
"loss": 1.0955,
"step": 181000
},
{
"epoch": 2.92,
"learning_rate": 1.2560627789684011e-06,
"loss": 1.0765,
"step": 181500
},
{
"epoch": 2.93,
"learning_rate": 1.1217819601776804e-06,
"loss": 1.0888,
"step": 182000
},
{
"epoch": 2.94,
"learning_rate": 9.875011413869599e-07,
"loss": 1.1032,
"step": 182500
},
{
"epoch": 2.95,
"learning_rate": 8.532203225962391e-07,
"loss": 1.087,
"step": 183000
},
{
"epoch": 2.96,
"learning_rate": 7.189395038055184e-07,
"loss": 1.1055,
"step": 183500
},
{
"epoch": 2.96,
"learning_rate": 5.846586850147977e-07,
"loss": 1.1,
"step": 184000
},
{
"epoch": 2.97,
"learning_rate": 4.503778662240771e-07,
"loss": 1.0934,
"step": 184500
},
{
"epoch": 2.98,
"learning_rate": 3.160970474333564e-07,
"loss": 1.0985,
"step": 185000
},
{
"epoch": 2.99,
"learning_rate": 1.818162286426358e-07,
"loss": 1.1007,
"step": 185500
},
{
"epoch": 3.0,
"learning_rate": 4.7535409851915116e-08,
"loss": 1.0886,
"step": 186000
},
{
"epoch": 3.0,
"step": 186177,
"total_flos": 4.045148042027827e+17,
"train_loss": 1.3697181449678724,
"train_runtime": 60540.4829,
"train_samples_per_second": 30.752,
"train_steps_per_second": 3.075
}
],
"max_steps": 186177,
"num_train_epochs": 3,
"total_flos": 4.045148042027827e+17,
"trial_name": null,
"trial_params": null
}