| { | |
| "best_metric": 0.1486110290722824, | |
| "best_model_checkpoint": "/bartabsa-reproduce/outputs/gpt22gpt2_42/checkpoint-18000", | |
| "epoch": 2.999832822513235, | |
| "eval_steps": 2000, | |
| "global_step": 26916, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05572582892170521, | |
| "grad_norm": 2.641549587249756, | |
| "learning_rate": 2.5e-05, | |
| "loss": 3.4506, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.11145165784341042, | |
| "grad_norm": 2.0933542251586914, | |
| "learning_rate": 5e-05, | |
| "loss": 2.9017, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.16717748676511562, | |
| "grad_norm": 1.5248066186904907, | |
| "learning_rate": 4.9035344960642076e-05, | |
| "loss": 2.7785, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.22290331568682084, | |
| "grad_norm": 1.6139370203018188, | |
| "learning_rate": 4.807068992128415e-05, | |
| "loss": 2.6006, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.22290331568682084, | |
| "eval_loss": 2.389116048812866, | |
| "eval_rouge1": 0.19037910692281135, | |
| "eval_rouge2": 0.04492833625707074, | |
| "eval_rougeL": 0.12708407141004266, | |
| "eval_rougeLsum": 0.1782021549831319, | |
| "eval_runtime": 6591.4182, | |
| "eval_samples_per_second": 2.028, | |
| "eval_steps_per_second": 0.127, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.27862914460852606, | |
| "grad_norm": 1.6594542264938354, | |
| "learning_rate": 4.7106034881926225e-05, | |
| "loss": 2.4813, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.33435497353023125, | |
| "grad_norm": 2.8870716094970703, | |
| "learning_rate": 4.61413798425683e-05, | |
| "loss": 2.3756, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.3900808024519365, | |
| "grad_norm": 1.8809521198272705, | |
| "learning_rate": 4.517672480321037e-05, | |
| "loss": 2.2931, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.4458066313736417, | |
| "grad_norm": 1.6554068326950073, | |
| "learning_rate": 4.421206976385245e-05, | |
| "loss": 2.1985, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.4458066313736417, | |
| "eval_loss": 2.0162405967712402, | |
| "eval_rouge1": 0.260762940287145, | |
| "eval_rouge2": 0.0874141148573599, | |
| "eval_rougeL": 0.1660542965132621, | |
| "eval_rougeLsum": 0.24517769329917644, | |
| "eval_runtime": 6637.6674, | |
| "eval_samples_per_second": 2.014, | |
| "eval_steps_per_second": 0.126, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.5015324602953469, | |
| "grad_norm": 1.5504436492919922, | |
| "learning_rate": 4.324741472449452e-05, | |
| "loss": 2.1302, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.5572582892170521, | |
| "grad_norm": 1.6397191286087036, | |
| "learning_rate": 4.2282759685136595e-05, | |
| "loss": 2.0752, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.6129841181387573, | |
| "grad_norm": 1.410211205482483, | |
| "learning_rate": 4.131810464577867e-05, | |
| "loss": 2.0196, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.6687099470604625, | |
| "grad_norm": 1.4629054069519043, | |
| "learning_rate": 4.035344960642074e-05, | |
| "loss": 1.9813, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.6687099470604625, | |
| "eval_loss": 1.8432687520980835, | |
| "eval_rouge1": 0.23899211300727613, | |
| "eval_rouge2": 0.07980074448958381, | |
| "eval_rougeL": 0.157016441992265, | |
| "eval_rougeLsum": 0.22665214645590098, | |
| "eval_runtime": 6625.3747, | |
| "eval_samples_per_second": 2.018, | |
| "eval_steps_per_second": 0.126, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.7244357759821677, | |
| "grad_norm": 1.630794644355774, | |
| "learning_rate": 3.938879456706282e-05, | |
| "loss": 1.9589, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.780161604903873, | |
| "grad_norm": 1.3507907390594482, | |
| "learning_rate": 3.84241395277049e-05, | |
| "loss": 1.9353, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.8358874338255782, | |
| "grad_norm": 1.4204024076461792, | |
| "learning_rate": 3.745948448834697e-05, | |
| "loss": 1.9153, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.8916132627472834, | |
| "grad_norm": 1.4473203420639038, | |
| "learning_rate": 3.6494829448989046e-05, | |
| "loss": 1.8954, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.8916132627472834, | |
| "eval_loss": 1.765182614326477, | |
| "eval_rouge1": 0.2693508099975971, | |
| "eval_rouge2": 0.09681331195925905, | |
| "eval_rougeL": 0.17175063229741494, | |
| "eval_rougeLsum": 0.25328693268916047, | |
| "eval_runtime": 6661.7822, | |
| "eval_samples_per_second": 2.007, | |
| "eval_steps_per_second": 0.125, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.9473390916689886, | |
| "grad_norm": 1.2767947912216187, | |
| "learning_rate": 3.553017440963112e-05, | |
| "loss": 1.8768, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.0030649205906939, | |
| "grad_norm": 1.5864351987838745, | |
| "learning_rate": 3.4565519370273194e-05, | |
| "loss": 1.8498, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.058790749512399, | |
| "grad_norm": 1.3982937335968018, | |
| "learning_rate": 3.360086433091527e-05, | |
| "loss": 1.6007, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.1145165784341042, | |
| "grad_norm": 1.3697761297225952, | |
| "learning_rate": 3.263620929155734e-05, | |
| "loss": 1.5988, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.1145165784341042, | |
| "eval_loss": 1.7400025129318237, | |
| "eval_rouge1": 0.3181180638016936, | |
| "eval_rouge2": 0.12747387109506586, | |
| "eval_rougeL": 0.1992092848661834, | |
| "eval_rougeLsum": 0.29982989222707246, | |
| "eval_runtime": 6684.9846, | |
| "eval_samples_per_second": 2.0, | |
| "eval_steps_per_second": 0.125, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.1702424073558095, | |
| "grad_norm": 1.364120364189148, | |
| "learning_rate": 3.1671554252199416e-05, | |
| "loss": 1.6006, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.2259682362775146, | |
| "grad_norm": 1.3193862438201904, | |
| "learning_rate": 3.070689921284149e-05, | |
| "loss": 1.6012, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.28169406519922, | |
| "grad_norm": 1.37600576877594, | |
| "learning_rate": 2.9742244173483564e-05, | |
| "loss": 1.5891, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.337419894120925, | |
| "grad_norm": 1.2682479619979858, | |
| "learning_rate": 2.8777589134125638e-05, | |
| "loss": 1.5897, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.337419894120925, | |
| "eval_loss": 1.7119464874267578, | |
| "eval_rouge1": 0.3291626272267019, | |
| "eval_rouge2": 0.13506936924825233, | |
| "eval_rougeL": 0.20493611801658168, | |
| "eval_rougeLsum": 0.31072127284396944, | |
| "eval_runtime": 6702.6268, | |
| "eval_samples_per_second": 1.994, | |
| "eval_steps_per_second": 0.125, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.3931457230426303, | |
| "grad_norm": 1.2508635520935059, | |
| "learning_rate": 2.7812934094767712e-05, | |
| "loss": 1.5867, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.4488715519643356, | |
| "grad_norm": 1.358458399772644, | |
| "learning_rate": 2.6848279055409786e-05, | |
| "loss": 1.5949, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.5045973808860407, | |
| "grad_norm": 1.41256582736969, | |
| "learning_rate": 2.588362401605186e-05, | |
| "loss": 1.5815, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.5603232098077457, | |
| "grad_norm": 1.3600412607192993, | |
| "learning_rate": 2.4918968976693934e-05, | |
| "loss": 1.5809, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.5603232098077457, | |
| "eval_loss": 1.692618489265442, | |
| "eval_rouge1": 0.3451533884469822, | |
| "eval_rouge2": 0.14511077647596418, | |
| "eval_rougeL": 0.2141526672382322, | |
| "eval_rougeLsum": 0.3261614308302348, | |
| "eval_runtime": 6740.4842, | |
| "eval_samples_per_second": 1.983, | |
| "eval_steps_per_second": 0.124, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.616049038729451, | |
| "grad_norm": 1.4034714698791504, | |
| "learning_rate": 2.3954313937336008e-05, | |
| "loss": 1.5799, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.6717748676511563, | |
| "grad_norm": 1.2388148307800293, | |
| "learning_rate": 2.2989658897978082e-05, | |
| "loss": 1.57, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.7275006965728616, | |
| "grad_norm": 1.284287452697754, | |
| "learning_rate": 2.2025003858620156e-05, | |
| "loss": 1.5642, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 1.7832265254945667, | |
| "grad_norm": 1.2729047536849976, | |
| "learning_rate": 2.1060348819262234e-05, | |
| "loss": 1.575, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.7832265254945667, | |
| "eval_loss": 1.6679071187973022, | |
| "eval_rouge1": 0.34402110027684635, | |
| "eval_rouge2": 0.14521497939582595, | |
| "eval_rougeL": 0.2148820791778157, | |
| "eval_rougeLsum": 0.32557257755274044, | |
| "eval_runtime": 6725.0049, | |
| "eval_samples_per_second": 1.988, | |
| "eval_steps_per_second": 0.124, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.8389523544162718, | |
| "grad_norm": 1.3148972988128662, | |
| "learning_rate": 2.0095693779904308e-05, | |
| "loss": 1.5683, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 1.894678183337977, | |
| "grad_norm": 1.3657242059707642, | |
| "learning_rate": 1.9131038740546382e-05, | |
| "loss": 1.556, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.9504040122596824, | |
| "grad_norm": 1.3399959802627563, | |
| "learning_rate": 1.8166383701188456e-05, | |
| "loss": 1.5561, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 2.0061298411813877, | |
| "grad_norm": 1.3426917791366577, | |
| "learning_rate": 1.720172866183053e-05, | |
| "loss": 1.5302, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.0061298411813877, | |
| "eval_loss": 1.686989188194275, | |
| "eval_rouge1": 0.35122976981430254, | |
| "eval_rouge2": 0.1486110290722824, | |
| "eval_rougeL": 0.21676569213257063, | |
| "eval_rougeLsum": 0.33162076027763865, | |
| "eval_runtime": 6697.8554, | |
| "eval_samples_per_second": 1.996, | |
| "eval_steps_per_second": 0.125, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.0618556701030926, | |
| "grad_norm": 1.3691883087158203, | |
| "learning_rate": 1.6237073622472604e-05, | |
| "loss": 1.2697, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 2.117581499024798, | |
| "grad_norm": 1.5133634805679321, | |
| "learning_rate": 1.5272418583114678e-05, | |
| "loss": 1.2697, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 2.173307327946503, | |
| "grad_norm": 1.532254934310913, | |
| "learning_rate": 1.4307763543756752e-05, | |
| "loss": 1.2763, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 2.2290331568682085, | |
| "grad_norm": 1.5224040746688843, | |
| "learning_rate": 1.3343108504398828e-05, | |
| "loss": 1.2726, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.2290331568682085, | |
| "eval_loss": 1.700244665145874, | |
| "eval_rouge1": 0.34844309743022917, | |
| "eval_rouge2": 0.14602958441027564, | |
| "eval_rougeL": 0.21492715808201202, | |
| "eval_rougeLsum": 0.3288645676067079, | |
| "eval_runtime": 6694.4601, | |
| "eval_samples_per_second": 1.997, | |
| "eval_steps_per_second": 0.125, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.2847589857899138, | |
| "grad_norm": 1.3866724967956543, | |
| "learning_rate": 1.2378453465040902e-05, | |
| "loss": 1.267, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 2.340484814711619, | |
| "grad_norm": 1.449537754058838, | |
| "learning_rate": 1.1413798425682977e-05, | |
| "loss": 1.2713, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 2.396210643633324, | |
| "grad_norm": 1.5382102727890015, | |
| "learning_rate": 1.0449143386325052e-05, | |
| "loss": 1.2661, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 2.4519364725550292, | |
| "grad_norm": 1.4987465143203735, | |
| "learning_rate": 9.484488346967126e-06, | |
| "loss": 1.266, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 2.4519364725550292, | |
| "eval_loss": 1.696866750717163, | |
| "eval_rouge1": 0.34729120268114544, | |
| "eval_rouge2": 0.1461422440790512, | |
| "eval_rougeL": 0.21543359860015138, | |
| "eval_rougeLsum": 0.3279325869793922, | |
| "eval_runtime": 6702.5269, | |
| "eval_samples_per_second": 1.994, | |
| "eval_steps_per_second": 0.125, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 2.5076623014767345, | |
| "grad_norm": 1.320014238357544, | |
| "learning_rate": 8.5198333076092e-06, | |
| "loss": 1.2621, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 2.56338813039844, | |
| "grad_norm": 1.5119301080703735, | |
| "learning_rate": 7.5551782682512745e-06, | |
| "loss": 1.2599, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 2.6191139593201447, | |
| "grad_norm": 1.6007628440856934, | |
| "learning_rate": 6.5905232288933485e-06, | |
| "loss": 1.2645, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 2.67483978824185, | |
| "grad_norm": 1.4694104194641113, | |
| "learning_rate": 5.6258681895354226e-06, | |
| "loss": 1.2566, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 2.67483978824185, | |
| "eval_loss": 1.6878169775009155, | |
| "eval_rouge1": 0.3487035276268111, | |
| "eval_rouge2": 0.14689326081743093, | |
| "eval_rougeL": 0.21600752679935203, | |
| "eval_rougeLsum": 0.3296249182423762, | |
| "eval_runtime": 6708.9033, | |
| "eval_samples_per_second": 1.993, | |
| "eval_steps_per_second": 0.125, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 2.7305656171635553, | |
| "grad_norm": 1.402747631072998, | |
| "learning_rate": 4.661213150177497e-06, | |
| "loss": 1.2663, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 2.7862914460852606, | |
| "grad_norm": 1.5145317316055298, | |
| "learning_rate": 3.6965581108195706e-06, | |
| "loss": 1.2573, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 2.842017275006966, | |
| "grad_norm": 1.4094128608703613, | |
| "learning_rate": 2.7319030714616455e-06, | |
| "loss": 1.2584, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 2.897743103928671, | |
| "grad_norm": 1.3608603477478027, | |
| "learning_rate": 1.7672480321037198e-06, | |
| "loss": 1.2572, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 2.897743103928671, | |
| "eval_loss": 1.6841516494750977, | |
| "eval_rouge1": 0.3499620308080743, | |
| "eval_rouge2": 0.1477254043640691, | |
| "eval_rougeL": 0.21694655364952178, | |
| "eval_rougeLsum": 0.3305722112116853, | |
| "eval_runtime": 6698.116, | |
| "eval_samples_per_second": 1.996, | |
| "eval_steps_per_second": 0.125, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 2.953468932850376, | |
| "grad_norm": 1.4232847690582275, | |
| "learning_rate": 8.025929927457941e-07, | |
| "loss": 1.2569, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 2.999832822513235, | |
| "step": 26916, | |
| "total_flos": 4.3736503104621773e+18, | |
| "train_loss": 1.7030430710285587, | |
| "train_runtime": 114489.2489, | |
| "train_samples_per_second": 7.523, | |
| "train_steps_per_second": 0.235 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 26916, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 2000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.3736503104621773e+18, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |