{ "best_metric": null, "best_model_checkpoint": null, "epoch": 3.0, "global_step": 73632, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.02, "learning_rate": 4.9660473707083876e-05, "loss": 0.8682, "step": 500 }, { "epoch": 0.04, "learning_rate": 4.932094741416776e-05, "loss": 0.7273, "step": 1000 }, { "epoch": 0.06, "learning_rate": 4.898142112125163e-05, "loss": 0.6857, "step": 1500 }, { "epoch": 0.08, "learning_rate": 4.864189482833551e-05, "loss": 0.6453, "step": 2000 }, { "epoch": 0.1, "learning_rate": 4.8302368535419385e-05, "loss": 0.6283, "step": 2500 }, { "epoch": 0.12, "learning_rate": 4.7962842242503265e-05, "loss": 0.6197, "step": 3000 }, { "epoch": 0.14, "learning_rate": 4.762331594958714e-05, "loss": 0.6029, "step": 3500 }, { "epoch": 0.16, "learning_rate": 4.728378965667101e-05, "loss": 0.5786, "step": 4000 }, { "epoch": 0.18, "learning_rate": 4.694426336375489e-05, "loss": 0.595, "step": 4500 }, { "epoch": 0.2, "learning_rate": 4.660473707083877e-05, "loss": 0.5902, "step": 5000 }, { "epoch": 0.22, "learning_rate": 4.626521077792265e-05, "loss": 0.5888, "step": 5500 }, { "epoch": 0.24, "learning_rate": 4.592568448500652e-05, "loss": 0.5708, "step": 6000 }, { "epoch": 0.26, "learning_rate": 4.55861581920904e-05, "loss": 0.5667, "step": 6500 }, { "epoch": 0.29, "learning_rate": 4.5246631899174276e-05, "loss": 0.5807, "step": 7000 }, { "epoch": 0.31, "learning_rate": 4.490710560625815e-05, "loss": 0.5721, "step": 7500 }, { "epoch": 0.33, "learning_rate": 4.456757931334203e-05, "loss": 0.5549, "step": 8000 }, { "epoch": 0.35, "learning_rate": 4.4228053020425904e-05, "loss": 0.546, "step": 8500 }, { "epoch": 0.37, "learning_rate": 4.3888526727509784e-05, "loss": 0.5376, "step": 9000 }, { "epoch": 0.39, "learning_rate": 4.354900043459366e-05, "loss": 0.5533, "step": 9500 }, { "epoch": 0.41, "learning_rate": 4.320947414167754e-05, "loss": 0.5457, "step": 10000 }, { "epoch": 0.43, "learning_rate": 4.286994784876141e-05, "loss": 0.551, "step": 10500 }, { "epoch": 0.45, "learning_rate": 4.2530421555845286e-05, "loss": 0.545, "step": 11000 }, { "epoch": 0.47, "learning_rate": 4.219089526292917e-05, "loss": 0.539, "step": 11500 }, { "epoch": 0.49, "learning_rate": 4.185136897001304e-05, "loss": 0.5354, "step": 12000 }, { "epoch": 0.51, "learning_rate": 4.151184267709692e-05, "loss": 0.5251, "step": 12500 }, { "epoch": 0.53, "learning_rate": 4.1172316384180795e-05, "loss": 0.5493, "step": 13000 }, { "epoch": 0.55, "learning_rate": 4.0832790091264675e-05, "loss": 0.5293, "step": 13500 }, { "epoch": 0.57, "learning_rate": 4.049326379834855e-05, "loss": 0.5349, "step": 14000 }, { "epoch": 0.59, "learning_rate": 4.015373750543242e-05, "loss": 0.5141, "step": 14500 }, { "epoch": 0.61, "learning_rate": 3.98142112125163e-05, "loss": 0.536, "step": 15000 }, { "epoch": 0.63, "learning_rate": 3.947468491960018e-05, "loss": 0.5209, "step": 15500 }, { "epoch": 0.65, "learning_rate": 3.913515862668405e-05, "loss": 0.5218, "step": 16000 }, { "epoch": 0.67, "learning_rate": 3.879563233376793e-05, "loss": 0.5294, "step": 16500 }, { "epoch": 0.69, "learning_rate": 3.8456106040851805e-05, "loss": 0.5154, "step": 17000 }, { "epoch": 0.71, "learning_rate": 3.811657974793568e-05, "loss": 0.524, "step": 17500 }, { "epoch": 0.73, "learning_rate": 3.777705345501956e-05, "loss": 0.5032, "step": 18000 }, { "epoch": 0.75, "learning_rate": 3.743752716210343e-05, "loss": 0.5083, "step": 18500 }, { "epoch": 0.77, "learning_rate": 3.7098000869187314e-05, "loss": 0.5192, "step": 19000 }, { "epoch": 0.79, "learning_rate": 3.675847457627119e-05, "loss": 0.5035, "step": 19500 }, { "epoch": 0.81, "learning_rate": 3.641894828335506e-05, "loss": 0.4907, "step": 20000 }, { "epoch": 0.84, "learning_rate": 3.607942199043894e-05, "loss": 0.51, "step": 20500 }, { "epoch": 0.86, "learning_rate": 3.5739895697522816e-05, "loss": 0.5032, "step": 21000 }, { "epoch": 0.88, "learning_rate": 3.540036940460669e-05, "loss": 0.5003, "step": 21500 }, { "epoch": 0.9, "learning_rate": 3.506084311169057e-05, "loss": 0.5053, "step": 22000 }, { "epoch": 0.92, "learning_rate": 3.4721316818774444e-05, "loss": 0.5073, "step": 22500 }, { "epoch": 0.94, "learning_rate": 3.4381790525858324e-05, "loss": 0.4997, "step": 23000 }, { "epoch": 0.96, "learning_rate": 3.40422642329422e-05, "loss": 0.4906, "step": 23500 }, { "epoch": 0.98, "learning_rate": 3.370273794002607e-05, "loss": 0.4974, "step": 24000 }, { "epoch": 1.0, "learning_rate": 3.336321164710995e-05, "loss": 0.4949, "step": 24500 }, { "epoch": 1.02, "learning_rate": 3.3023685354193826e-05, "loss": 0.3855, "step": 25000 }, { "epoch": 1.04, "learning_rate": 3.2684159061277707e-05, "loss": 0.3692, "step": 25500 }, { "epoch": 1.06, "learning_rate": 3.234463276836158e-05, "loss": 0.3659, "step": 26000 }, { "epoch": 1.08, "learning_rate": 3.200510647544546e-05, "loss": 0.3748, "step": 26500 }, { "epoch": 1.1, "learning_rate": 3.1665580182529335e-05, "loss": 0.3547, "step": 27000 }, { "epoch": 1.12, "learning_rate": 3.132605388961321e-05, "loss": 0.3658, "step": 27500 }, { "epoch": 1.14, "learning_rate": 3.098652759669709e-05, "loss": 0.3764, "step": 28000 }, { "epoch": 1.16, "learning_rate": 3.064700130378096e-05, "loss": 0.3796, "step": 28500 }, { "epoch": 1.18, "learning_rate": 3.0307475010864843e-05, "loss": 0.3742, "step": 29000 }, { "epoch": 1.2, "learning_rate": 2.9967948717948717e-05, "loss": 0.3798, "step": 29500 }, { "epoch": 1.22, "learning_rate": 2.9628422425032598e-05, "loss": 0.3776, "step": 30000 }, { "epoch": 1.24, "learning_rate": 2.928889613211647e-05, "loss": 0.3689, "step": 30500 }, { "epoch": 1.26, "learning_rate": 2.8949369839200345e-05, "loss": 0.3627, "step": 31000 }, { "epoch": 1.28, "learning_rate": 2.8609843546284226e-05, "loss": 0.3686, "step": 31500 }, { "epoch": 1.3, "learning_rate": 2.82703172533681e-05, "loss": 0.3748, "step": 32000 }, { "epoch": 1.32, "learning_rate": 2.793079096045198e-05, "loss": 0.3773, "step": 32500 }, { "epoch": 1.34, "learning_rate": 2.7591264667535854e-05, "loss": 0.3747, "step": 33000 }, { "epoch": 1.36, "learning_rate": 2.7251738374619734e-05, "loss": 0.3846, "step": 33500 }, { "epoch": 1.39, "learning_rate": 2.6912212081703608e-05, "loss": 0.3792, "step": 34000 }, { "epoch": 1.41, "learning_rate": 2.6572685788787482e-05, "loss": 0.3594, "step": 34500 }, { "epoch": 1.43, "learning_rate": 2.6233159495871362e-05, "loss": 0.3869, "step": 35000 }, { "epoch": 1.45, "learning_rate": 2.5893633202955236e-05, "loss": 0.3674, "step": 35500 }, { "epoch": 1.47, "learning_rate": 2.5554106910039117e-05, "loss": 0.3642, "step": 36000 }, { "epoch": 1.49, "learning_rate": 2.521458061712299e-05, "loss": 0.3608, "step": 36500 }, { "epoch": 1.51, "learning_rate": 2.4875054324206867e-05, "loss": 0.3739, "step": 37000 }, { "epoch": 1.53, "learning_rate": 2.4535528031290745e-05, "loss": 0.3662, "step": 37500 }, { "epoch": 1.55, "learning_rate": 2.4196001738374622e-05, "loss": 0.3646, "step": 38000 }, { "epoch": 1.57, "learning_rate": 2.38564754454585e-05, "loss": 0.3811, "step": 38500 }, { "epoch": 1.59, "learning_rate": 2.3516949152542376e-05, "loss": 0.3568, "step": 39000 }, { "epoch": 1.61, "learning_rate": 2.317742285962625e-05, "loss": 0.3663, "step": 39500 }, { "epoch": 1.63, "learning_rate": 2.2837896566710127e-05, "loss": 0.3716, "step": 40000 }, { "epoch": 1.65, "learning_rate": 2.2498370273794004e-05, "loss": 0.3562, "step": 40500 }, { "epoch": 1.67, "learning_rate": 2.215884398087788e-05, "loss": 0.3769, "step": 41000 }, { "epoch": 1.69, "learning_rate": 2.181931768796176e-05, "loss": 0.3712, "step": 41500 }, { "epoch": 1.71, "learning_rate": 2.1479791395045636e-05, "loss": 0.3691, "step": 42000 }, { "epoch": 1.73, "learning_rate": 2.1140265102129513e-05, "loss": 0.369, "step": 42500 }, { "epoch": 1.75, "learning_rate": 2.0800738809213386e-05, "loss": 0.3644, "step": 43000 }, { "epoch": 1.77, "learning_rate": 2.0461212516297264e-05, "loss": 0.3611, "step": 43500 }, { "epoch": 1.79, "learning_rate": 2.012168622338114e-05, "loss": 0.357, "step": 44000 }, { "epoch": 1.81, "learning_rate": 1.9782159930465018e-05, "loss": 0.377, "step": 44500 }, { "epoch": 1.83, "learning_rate": 1.944263363754889e-05, "loss": 0.3562, "step": 45000 }, { "epoch": 1.85, "learning_rate": 1.910310734463277e-05, "loss": 0.3466, "step": 45500 }, { "epoch": 1.87, "learning_rate": 1.8763581051716646e-05, "loss": 0.3519, "step": 46000 }, { "epoch": 1.89, "learning_rate": 1.842405475880052e-05, "loss": 0.3514, "step": 46500 }, { "epoch": 1.91, "learning_rate": 1.8084528465884397e-05, "loss": 0.3498, "step": 47000 }, { "epoch": 1.94, "learning_rate": 1.7745002172968274e-05, "loss": 0.3575, "step": 47500 }, { "epoch": 1.96, "learning_rate": 1.740547588005215e-05, "loss": 0.3398, "step": 48000 }, { "epoch": 1.98, "learning_rate": 1.706594958713603e-05, "loss": 0.3515, "step": 48500 }, { "epoch": 2.0, "learning_rate": 1.6726423294219905e-05, "loss": 0.3594, "step": 49000 }, { "epoch": 2.02, "learning_rate": 1.6386897001303783e-05, "loss": 0.2482, "step": 49500 }, { "epoch": 2.04, "learning_rate": 1.6047370708387656e-05, "loss": 0.2321, "step": 50000 }, { "epoch": 2.06, "learning_rate": 1.5707844415471534e-05, "loss": 0.2207, "step": 50500 }, { "epoch": 2.08, "learning_rate": 1.536831812255541e-05, "loss": 0.2141, "step": 51000 }, { "epoch": 2.1, "learning_rate": 1.5028791829639288e-05, "loss": 0.2253, "step": 51500 }, { "epoch": 2.12, "learning_rate": 1.4689265536723165e-05, "loss": 0.2415, "step": 52000 }, { "epoch": 2.14, "learning_rate": 1.4349739243807042e-05, "loss": 0.2079, "step": 52500 }, { "epoch": 2.16, "learning_rate": 1.4010212950890916e-05, "loss": 0.2156, "step": 53000 }, { "epoch": 2.18, "learning_rate": 1.3670686657974793e-05, "loss": 0.216, "step": 53500 }, { "epoch": 2.2, "learning_rate": 1.333116036505867e-05, "loss": 0.2245, "step": 54000 }, { "epoch": 2.22, "learning_rate": 1.2991634072142547e-05, "loss": 0.2209, "step": 54500 }, { "epoch": 2.24, "learning_rate": 1.2652107779226425e-05, "loss": 0.2121, "step": 55000 }, { "epoch": 2.26, "learning_rate": 1.23125814863103e-05, "loss": 0.2077, "step": 55500 }, { "epoch": 2.28, "learning_rate": 1.1973055193394177e-05, "loss": 0.2181, "step": 56000 }, { "epoch": 2.3, "learning_rate": 1.1633528900478054e-05, "loss": 0.221, "step": 56500 }, { "epoch": 2.32, "learning_rate": 1.1294002607561931e-05, "loss": 0.2275, "step": 57000 }, { "epoch": 2.34, "learning_rate": 1.0954476314645807e-05, "loss": 0.2258, "step": 57500 }, { "epoch": 2.36, "learning_rate": 1.0614950021729684e-05, "loss": 0.2246, "step": 58000 }, { "epoch": 2.38, "learning_rate": 1.027542372881356e-05, "loss": 0.2235, "step": 58500 }, { "epoch": 2.4, "learning_rate": 9.935897435897435e-06, "loss": 0.2123, "step": 59000 }, { "epoch": 2.42, "learning_rate": 9.596371142981312e-06, "loss": 0.2015, "step": 59500 }, { "epoch": 2.44, "learning_rate": 9.25684485006519e-06, "loss": 0.2204, "step": 60000 }, { "epoch": 2.46, "learning_rate": 8.917318557149066e-06, "loss": 0.2243, "step": 60500 }, { "epoch": 2.49, "learning_rate": 8.577792264232942e-06, "loss": 0.2134, "step": 61000 }, { "epoch": 2.51, "learning_rate": 8.238265971316819e-06, "loss": 0.2121, "step": 61500 }, { "epoch": 2.53, "learning_rate": 7.898739678400696e-06, "loss": 0.2147, "step": 62000 }, { "epoch": 2.55, "learning_rate": 7.559213385484572e-06, "loss": 0.2279, "step": 62500 }, { "epoch": 2.57, "learning_rate": 7.219687092568449e-06, "loss": 0.2051, "step": 63000 }, { "epoch": 2.59, "learning_rate": 6.880160799652326e-06, "loss": 0.2262, "step": 63500 }, { "epoch": 2.61, "learning_rate": 6.540634506736203e-06, "loss": 0.2125, "step": 64000 }, { "epoch": 2.63, "learning_rate": 6.2011082138200785e-06, "loss": 0.2235, "step": 64500 }, { "epoch": 2.65, "learning_rate": 5.861581920903955e-06, "loss": 0.2113, "step": 65000 }, { "epoch": 2.67, "learning_rate": 5.522055627987831e-06, "loss": 0.2281, "step": 65500 }, { "epoch": 2.69, "learning_rate": 5.182529335071708e-06, "loss": 0.203, "step": 66000 }, { "epoch": 2.71, "learning_rate": 4.8430030421555846e-06, "loss": 0.2096, "step": 66500 }, { "epoch": 2.73, "learning_rate": 4.503476749239461e-06, "loss": 0.2076, "step": 67000 }, { "epoch": 2.75, "learning_rate": 4.163950456323338e-06, "loss": 0.209, "step": 67500 }, { "epoch": 2.77, "learning_rate": 3.824424163407214e-06, "loss": 0.2124, "step": 68000 }, { "epoch": 2.79, "learning_rate": 3.484897870491091e-06, "loss": 0.2091, "step": 68500 }, { "epoch": 2.81, "learning_rate": 3.1453715775749674e-06, "loss": 0.2147, "step": 69000 }, { "epoch": 2.83, "learning_rate": 2.805845284658844e-06, "loss": 0.2153, "step": 69500 }, { "epoch": 2.85, "learning_rate": 2.466318991742721e-06, "loss": 0.2221, "step": 70000 }, { "epoch": 2.87, "learning_rate": 2.126792698826597e-06, "loss": 0.225, "step": 70500 }, { "epoch": 2.89, "learning_rate": 1.7872664059104738e-06, "loss": 0.2068, "step": 71000 }, { "epoch": 2.91, "learning_rate": 1.4477401129943503e-06, "loss": 0.2147, "step": 71500 }, { "epoch": 2.93, "learning_rate": 1.108213820078227e-06, "loss": 0.2133, "step": 72000 }, { "epoch": 2.95, "learning_rate": 7.686875271621035e-07, "loss": 0.209, "step": 72500 }, { "epoch": 2.97, "learning_rate": 4.2916123424598005e-07, "loss": 0.2092, "step": 73000 }, { "epoch": 2.99, "learning_rate": 8.963494132985658e-08, "loss": 0.2103, "step": 73500 }, { "epoch": 3.0, "step": 73632, "total_flos": 7.74938740264658e+16, "train_loss": 0.3792748264984585, "train_runtime": 5994.6559, "train_samples_per_second": 196.526, "train_steps_per_second": 12.283 } ], "max_steps": 73632, "num_train_epochs": 3, "total_flos": 7.74938740264658e+16, "trial_name": null, "trial_params": null }