| { | |
| "best_metric": 0.9075653971518505, | |
| "best_model_checkpoint": "output/fine_tuned/roberta-large/QQP/checkpoint-22744", | |
| "epoch": 6.0, | |
| "eval_steps": 500, | |
| "global_step": 34116, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.08793527963418923, | |
| "grad_norm": 8.288018226623535, | |
| "learning_rate": 1.970688240121937e-05, | |
| "loss": 0.3997, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.17587055926837847, | |
| "grad_norm": 4.3138747215271, | |
| "learning_rate": 1.9413764802438738e-05, | |
| "loss": 0.3023, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.2638058389025677, | |
| "grad_norm": 6.374757289886475, | |
| "learning_rate": 1.9120647203658108e-05, | |
| "loss": 0.2839, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.35174111853675694, | |
| "grad_norm": 6.788768768310547, | |
| "learning_rate": 1.8827529604877478e-05, | |
| "loss": 0.2854, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.43967639817094617, | |
| "grad_norm": 16.165164947509766, | |
| "learning_rate": 1.8534412006096848e-05, | |
| "loss": 0.2674, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.5276116778051354, | |
| "grad_norm": 8.882967948913574, | |
| "learning_rate": 1.8241294407316218e-05, | |
| "loss": 0.2647, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.6155469574393246, | |
| "grad_norm": 7.7459869384765625, | |
| "learning_rate": 1.7948176808535588e-05, | |
| "loss": 0.2544, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.7034822370735139, | |
| "grad_norm": 7.243208885192871, | |
| "learning_rate": 1.7655059209754954e-05, | |
| "loss": 0.2541, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.7914175167077031, | |
| "grad_norm": 12.087061882019043, | |
| "learning_rate": 1.7361941610974324e-05, | |
| "loss": 0.2485, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.8793527963418923, | |
| "grad_norm": 5.829019546508789, | |
| "learning_rate": 1.7068824012193694e-05, | |
| "loss": 0.2484, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.9672880759760816, | |
| "grad_norm": 8.462845802307129, | |
| "learning_rate": 1.6775706413413064e-05, | |
| "loss": 0.2392, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9049220875587435, | |
| "eval_combined_score": 0.8867021824135581, | |
| "eval_f1": 0.8684822772683727, | |
| "eval_loss": 0.24716497957706451, | |
| "eval_runtime": 163.5025, | |
| "eval_samples_per_second": 247.274, | |
| "eval_steps_per_second": 30.911, | |
| "step": 5686 | |
| }, | |
| { | |
| "epoch": 1.0552233556102708, | |
| "grad_norm": 6.706808090209961, | |
| "learning_rate": 1.6482588814632434e-05, | |
| "loss": 0.2102, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.14315863524446, | |
| "grad_norm": 5.22786808013916, | |
| "learning_rate": 1.6189471215851803e-05, | |
| "loss": 0.1941, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.2310939148786493, | |
| "grad_norm": 6.846224784851074, | |
| "learning_rate": 1.589635361707117e-05, | |
| "loss": 0.1881, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.3190291945128385, | |
| "grad_norm": 7.269466876983643, | |
| "learning_rate": 1.560323601829054e-05, | |
| "loss": 0.183, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.4069644741470277, | |
| "grad_norm": 5.1561665534973145, | |
| "learning_rate": 1.531011841950991e-05, | |
| "loss": 0.1914, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.494899753781217, | |
| "grad_norm": 8.569211959838867, | |
| "learning_rate": 1.5017000820729278e-05, | |
| "loss": 0.1881, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.5828350334154062, | |
| "grad_norm": 7.138071060180664, | |
| "learning_rate": 1.4723883221948648e-05, | |
| "loss": 0.187, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.6707703130495954, | |
| "grad_norm": 5.725489139556885, | |
| "learning_rate": 1.4430765623168018e-05, | |
| "loss": 0.1905, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.7587055926837847, | |
| "grad_norm": 5.3855180740356445, | |
| "learning_rate": 1.4137648024387384e-05, | |
| "loss": 0.1859, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.8466408723179741, | |
| "grad_norm": 6.059995651245117, | |
| "learning_rate": 1.3844530425606754e-05, | |
| "loss": 0.186, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.9345761519521631, | |
| "grad_norm": 7.908290863037109, | |
| "learning_rate": 1.3551412826826124e-05, | |
| "loss": 0.1884, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.908904279000742, | |
| "eval_combined_score": 0.896049297852469, | |
| "eval_f1": 0.8831943167041959, | |
| "eval_loss": 0.23078934848308563, | |
| "eval_runtime": 163.6895, | |
| "eval_samples_per_second": 246.992, | |
| "eval_steps_per_second": 30.876, | |
| "step": 11372 | |
| }, | |
| { | |
| "epoch": 2.0225114315863526, | |
| "grad_norm": 7.034117698669434, | |
| "learning_rate": 1.3258295228045492e-05, | |
| "loss": 0.1764, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 2.1104467112205416, | |
| "grad_norm": 6.000309467315674, | |
| "learning_rate": 1.2965177629264862e-05, | |
| "loss": 0.1409, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.198381990854731, | |
| "grad_norm": 5.445409297943115, | |
| "learning_rate": 1.2672060030484232e-05, | |
| "loss": 0.1458, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 2.28631727048892, | |
| "grad_norm": 11.053699493408203, | |
| "learning_rate": 1.23789424317036e-05, | |
| "loss": 0.1408, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 2.3742525501231095, | |
| "grad_norm": 12.287221908569336, | |
| "learning_rate": 1.208582483292297e-05, | |
| "loss": 0.1465, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 2.4621878297572986, | |
| "grad_norm": 2.536963939666748, | |
| "learning_rate": 1.179270723414234e-05, | |
| "loss": 0.1398, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 2.550123109391488, | |
| "grad_norm": 3.656428813934326, | |
| "learning_rate": 1.1499589635361708e-05, | |
| "loss": 0.1391, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 2.638058389025677, | |
| "grad_norm": 4.2873053550720215, | |
| "learning_rate": 1.1206472036581078e-05, | |
| "loss": 0.1407, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.7259936686598665, | |
| "grad_norm": 8.166319847106934, | |
| "learning_rate": 1.0913354437800448e-05, | |
| "loss": 0.1401, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 2.8139289482940555, | |
| "grad_norm": 11.278782844543457, | |
| "learning_rate": 1.0620236839019815e-05, | |
| "loss": 0.1435, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.901864227928245, | |
| "grad_norm": 9.329524993896484, | |
| "learning_rate": 1.0327119240239184e-05, | |
| "loss": 0.1405, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 2.989799507562434, | |
| "grad_norm": 7.3363356590271, | |
| "learning_rate": 1.0034001641458554e-05, | |
| "loss": 0.1351, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9179569626514964, | |
| "eval_combined_score": 0.9036969907087687, | |
| "eval_f1": 0.8894370187660411, | |
| "eval_loss": 0.23165978491306305, | |
| "eval_runtime": 163.4119, | |
| "eval_samples_per_second": 247.412, | |
| "eval_steps_per_second": 30.928, | |
| "step": 17058 | |
| }, | |
| { | |
| "epoch": 3.0777347871966234, | |
| "grad_norm": 3.7927825450897217, | |
| "learning_rate": 9.740884042677923e-06, | |
| "loss": 0.1066, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 3.1656700668308124, | |
| "grad_norm": 28.014644622802734, | |
| "learning_rate": 9.447766443897292e-06, | |
| "loss": 0.1089, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 3.253605346465002, | |
| "grad_norm": 3.5056657791137695, | |
| "learning_rate": 9.154648845116662e-06, | |
| "loss": 0.1085, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 3.341540626099191, | |
| "grad_norm": 3.618999481201172, | |
| "learning_rate": 8.86153124633603e-06, | |
| "loss": 0.1114, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 3.4294759057333803, | |
| "grad_norm": 7.726201057434082, | |
| "learning_rate": 8.5684136475554e-06, | |
| "loss": 0.1096, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 3.5174111853675694, | |
| "grad_norm": 10.587724685668945, | |
| "learning_rate": 8.275296048774769e-06, | |
| "loss": 0.1064, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 3.605346465001759, | |
| "grad_norm": 6.7054948806762695, | |
| "learning_rate": 7.982178449994138e-06, | |
| "loss": 0.1067, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 3.693281744635948, | |
| "grad_norm": 5.118494987487793, | |
| "learning_rate": 7.689060851213508e-06, | |
| "loss": 0.1067, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 3.7812170242701373, | |
| "grad_norm": 5.718533039093018, | |
| "learning_rate": 7.395943252432877e-06, | |
| "loss": 0.107, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 3.8691523039043263, | |
| "grad_norm": 3.717242479324341, | |
| "learning_rate": 7.102825653652246e-06, | |
| "loss": 0.105, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 3.9570875835385158, | |
| "grad_norm": 5.7518439292907715, | |
| "learning_rate": 6.8097080548716155e-06, | |
| "loss": 0.1051, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9207024486767252, | |
| "eval_combined_score": 0.9075653971518505, | |
| "eval_f1": 0.8944283456269757, | |
| "eval_loss": 0.27043241262435913, | |
| "eval_runtime": 165.1459, | |
| "eval_samples_per_second": 244.814, | |
| "eval_steps_per_second": 30.603, | |
| "step": 22744 | |
| }, | |
| { | |
| "epoch": 4.045022863172705, | |
| "grad_norm": 8.56693172454834, | |
| "learning_rate": 6.5165904560909846e-06, | |
| "loss": 0.0944, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 4.132958142806894, | |
| "grad_norm": 0.9398753046989441, | |
| "learning_rate": 6.223472857310353e-06, | |
| "loss": 0.0836, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 4.220893422441083, | |
| "grad_norm": 11.362903594970703, | |
| "learning_rate": 5.930355258529723e-06, | |
| "loss": 0.0786, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 4.308828702075273, | |
| "grad_norm": 10.553667068481445, | |
| "learning_rate": 5.637237659749092e-06, | |
| "loss": 0.0805, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 4.396763981709462, | |
| "grad_norm": 2.639619827270508, | |
| "learning_rate": 5.344120060968461e-06, | |
| "loss": 0.0779, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 4.484699261343651, | |
| "grad_norm": 4.3451008796691895, | |
| "learning_rate": 5.051002462187831e-06, | |
| "loss": 0.0778, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 4.57263454097784, | |
| "grad_norm": 1.3628908395767212, | |
| "learning_rate": 4.7578848634072e-06, | |
| "loss": 0.0794, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 4.66056982061203, | |
| "grad_norm": 14.94767951965332, | |
| "learning_rate": 4.464767264626568e-06, | |
| "loss": 0.08, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 4.748505100246219, | |
| "grad_norm": 3.760082721710205, | |
| "learning_rate": 4.171649665845938e-06, | |
| "loss": 0.0785, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 4.836440379880408, | |
| "grad_norm": 9.946064949035645, | |
| "learning_rate": 3.878532067065307e-06, | |
| "loss": 0.0809, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 4.924375659514597, | |
| "grad_norm": 16.117393493652344, | |
| "learning_rate": 3.585414468284676e-06, | |
| "loss": 0.0794, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9194410091516201, | |
| "eval_combined_score": 0.9052938477127372, | |
| "eval_f1": 0.8911466862738545, | |
| "eval_loss": 0.32720956206321716, | |
| "eval_runtime": 162.6906, | |
| "eval_samples_per_second": 248.508, | |
| "eval_steps_per_second": 31.065, | |
| "step": 28430 | |
| }, | |
| { | |
| "epoch": 5.012310939148787, | |
| "grad_norm": 27.78730010986328, | |
| "learning_rate": 3.2922968695040454e-06, | |
| "loss": 0.0743, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 5.100246218782976, | |
| "grad_norm": 8.635354995727539, | |
| "learning_rate": 2.999179270723415e-06, | |
| "loss": 0.0599, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 5.188181498417165, | |
| "grad_norm": 11.963053703308105, | |
| "learning_rate": 2.7060616719427835e-06, | |
| "loss": 0.0619, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 5.276116778051354, | |
| "grad_norm": 3.372576951980591, | |
| "learning_rate": 2.412944073162153e-06, | |
| "loss": 0.0606, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 5.3640520576855435, | |
| "grad_norm": 6.7732343673706055, | |
| "learning_rate": 2.119826474381522e-06, | |
| "loss": 0.0616, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 5.451987337319733, | |
| "grad_norm": 11.202216148376465, | |
| "learning_rate": 1.8267088756008912e-06, | |
| "loss": 0.0662, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 5.5399226169539215, | |
| "grad_norm": 3.270718574523926, | |
| "learning_rate": 1.5335912768202605e-06, | |
| "loss": 0.0543, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 5.627857896588111, | |
| "grad_norm": 10.084086418151855, | |
| "learning_rate": 1.2404736780396295e-06, | |
| "loss": 0.0578, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 5.7157931762223, | |
| "grad_norm": 212.55088806152344, | |
| "learning_rate": 9.473560792589988e-07, | |
| "loss": 0.0596, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 5.80372845585649, | |
| "grad_norm": 0.6594523787498474, | |
| "learning_rate": 6.542384804783681e-07, | |
| "loss": 0.0589, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 5.891663735490679, | |
| "grad_norm": 11.235882759094238, | |
| "learning_rate": 3.611208816977372e-07, | |
| "loss": 0.0615, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 5.979599015124868, | |
| "grad_norm": 18.778226852416992, | |
| "learning_rate": 6.800328291710635e-08, | |
| "loss": 0.0564, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9206777145683898, | |
| "eval_combined_score": 0.9073393530700153, | |
| "eval_f1": 0.894000991571641, | |
| "eval_loss": 0.35158464312553406, | |
| "eval_runtime": 162.3775, | |
| "eval_samples_per_second": 248.988, | |
| "eval_steps_per_second": 31.125, | |
| "step": 34116 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "step": 34116, | |
| "total_flos": 5.086192482321592e+17, | |
| "train_loss": 0.14204222600531893, | |
| "train_runtime": 24723.4908, | |
| "train_samples_per_second": 88.3, | |
| "train_steps_per_second": 1.38 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 34116, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 6, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.086192482321592e+17, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |