| { | |
| "best_metric": 0.09976539827685843, | |
| "best_model_checkpoint": "/bartabsa-reproduce/outputs/gpt22gpt2_42_std_percentile/checkpoint-16000", | |
| "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": 2.3220622539520264, | |
| "learning_rate": 2.5e-05, | |
| "loss": 3.4893, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.11145165784341042, | |
| "grad_norm": 2.527012586593628, | |
| "learning_rate": 5e-05, | |
| "loss": 2.9952, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.16717748676511562, | |
| "grad_norm": 1.4452062845230103, | |
| "learning_rate": 4.9035344960642076e-05, | |
| "loss": 2.8253, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.22290331568682084, | |
| "grad_norm": 1.5733131170272827, | |
| "learning_rate": 4.807068992128415e-05, | |
| "loss": 2.7029, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.22290331568682084, | |
| "eval_loss": 2.542654514312744, | |
| "eval_rouge1": 0.24369185533364146, | |
| "eval_rouge2": 0.053150811665838404, | |
| "eval_rougeL": 0.1468615883810676, | |
| "eval_rougeLsum": 0.22845947792903248, | |
| "eval_runtime": 6410.9465, | |
| "eval_samples_per_second": 2.085, | |
| "eval_steps_per_second": 0.13, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.27862914460852606, | |
| "grad_norm": 1.733468770980835, | |
| "learning_rate": 4.7106034881926225e-05, | |
| "loss": 2.6421, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.33435497353023125, | |
| "grad_norm": 1.6630736589431763, | |
| "learning_rate": 4.61413798425683e-05, | |
| "loss": 2.6013, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.3900808024519365, | |
| "grad_norm": 3.9945011138916016, | |
| "learning_rate": 4.517672480321037e-05, | |
| "loss": 2.5585, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.4458066313736417, | |
| "grad_norm": 1.595192313194275, | |
| "learning_rate": 4.421206976385245e-05, | |
| "loss": 2.5218, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.4458066313736417, | |
| "eval_loss": 2.3777079582214355, | |
| "eval_rouge1": 0.26361751739528855, | |
| "eval_rouge2": 0.06605883769296508, | |
| "eval_rougeL": 0.15780235954372396, | |
| "eval_rougeLsum": 0.24547300255336824, | |
| "eval_runtime": 6409.59, | |
| "eval_samples_per_second": 2.086, | |
| "eval_steps_per_second": 0.13, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.5015324602953469, | |
| "grad_norm": 1.5933812856674194, | |
| "learning_rate": 4.324741472449452e-05, | |
| "loss": 2.4992, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.5572582892170521, | |
| "grad_norm": 1.5219066143035889, | |
| "learning_rate": 4.2282759685136595e-05, | |
| "loss": 2.4691, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.6129841181387573, | |
| "grad_norm": 1.4694565534591675, | |
| "learning_rate": 4.131810464577867e-05, | |
| "loss": 2.4356, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.6687099470604625, | |
| "grad_norm": 1.9962424039840698, | |
| "learning_rate": 4.035344960642074e-05, | |
| "loss": 2.4141, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.6687099470604625, | |
| "eval_loss": 2.2899107933044434, | |
| "eval_rouge1": 0.2682906048506787, | |
| "eval_rouge2": 0.06950238260179739, | |
| "eval_rougeL": 0.1587441828998173, | |
| "eval_rougeLsum": 0.25060189239845465, | |
| "eval_runtime": 6396.129, | |
| "eval_samples_per_second": 2.09, | |
| "eval_steps_per_second": 0.131, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.7244357759821677, | |
| "grad_norm": 6.416402816772461, | |
| "learning_rate": 3.938879456706282e-05, | |
| "loss": 2.3945, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.780161604903873, | |
| "grad_norm": 1.5234980583190918, | |
| "learning_rate": 3.84241395277049e-05, | |
| "loss": 2.3738, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.8358874338255782, | |
| "grad_norm": 4.883070468902588, | |
| "learning_rate": 3.745948448834697e-05, | |
| "loss": 2.3584, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.8916132627472834, | |
| "grad_norm": 1.8043371438980103, | |
| "learning_rate": 3.6494829448989046e-05, | |
| "loss": 2.3452, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.8916132627472834, | |
| "eval_loss": 2.219862937927246, | |
| "eval_rouge1": 0.2887851100955251, | |
| "eval_rouge2": 0.08135324671358818, | |
| "eval_rougeL": 0.17143312551180023, | |
| "eval_rougeLsum": 0.2693751508675857, | |
| "eval_runtime": 6386.9905, | |
| "eval_samples_per_second": 2.093, | |
| "eval_steps_per_second": 0.131, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.9473390916689886, | |
| "grad_norm": 2.5791819095611572, | |
| "learning_rate": 3.553017440963112e-05, | |
| "loss": 2.325, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.0030649205906939, | |
| "grad_norm": 1.7074276208877563, | |
| "learning_rate": 3.4565519370273194e-05, | |
| "loss": 2.2897, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.058790749512399, | |
| "grad_norm": 1.7298952341079712, | |
| "learning_rate": 3.360086433091527e-05, | |
| "loss": 2.0062, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.1145165784341042, | |
| "grad_norm": 1.575429916381836, | |
| "learning_rate": 3.263620929155734e-05, | |
| "loss": 2.0037, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.1145165784341042, | |
| "eval_loss": 2.1816983222961426, | |
| "eval_rouge1": 0.29842511392124504, | |
| "eval_rouge2": 0.08508044071327496, | |
| "eval_rougeL": 0.17288333694802113, | |
| "eval_rougeLsum": 0.2771806390068811, | |
| "eval_runtime": 6383.653, | |
| "eval_samples_per_second": 2.094, | |
| "eval_steps_per_second": 0.131, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.1702424073558095, | |
| "grad_norm": 1.666231632232666, | |
| "learning_rate": 3.1671554252199416e-05, | |
| "loss": 2.0088, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.2259682362775146, | |
| "grad_norm": 1.6241954565048218, | |
| "learning_rate": 3.070689921284149e-05, | |
| "loss": 2.0029, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.28169406519922, | |
| "grad_norm": 1.9751548767089844, | |
| "learning_rate": 2.9742244173483564e-05, | |
| "loss": 1.9862, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.337419894120925, | |
| "grad_norm": 13.691629409790039, | |
| "learning_rate": 2.8777589134125638e-05, | |
| "loss": 1.9905, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.337419894120925, | |
| "eval_loss": 2.1423134803771973, | |
| "eval_rouge1": 0.29073567485913165, | |
| "eval_rouge2": 0.08227824922946644, | |
| "eval_rougeL": 0.17132293154574824, | |
| "eval_rougeLsum": 0.2718749763727648, | |
| "eval_runtime": 6368.6616, | |
| "eval_samples_per_second": 2.099, | |
| "eval_steps_per_second": 0.131, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.3931457230426303, | |
| "grad_norm": 1.6704634428024292, | |
| "learning_rate": 2.7812934094767712e-05, | |
| "loss": 1.9886, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.4488715519643356, | |
| "grad_norm": 2.2663302421569824, | |
| "learning_rate": 2.6848279055409786e-05, | |
| "loss": 1.9843, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.5045973808860407, | |
| "grad_norm": 1.9374725818634033, | |
| "learning_rate": 2.588362401605186e-05, | |
| "loss": 1.9773, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.5603232098077457, | |
| "grad_norm": 2.878082513809204, | |
| "learning_rate": 2.4918968976693934e-05, | |
| "loss": 1.9711, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.5603232098077457, | |
| "eval_loss": 2.0982916355133057, | |
| "eval_rouge1": 0.3054580501233143, | |
| "eval_rouge2": 0.09153227846352645, | |
| "eval_rougeL": 0.1785600815773608, | |
| "eval_rougeLsum": 0.2849821221895824, | |
| "eval_runtime": 6387.6311, | |
| "eval_samples_per_second": 2.093, | |
| "eval_steps_per_second": 0.131, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.616049038729451, | |
| "grad_norm": 3.682783365249634, | |
| "learning_rate": 2.3954313937336008e-05, | |
| "loss": 1.9642, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.6717748676511563, | |
| "grad_norm": 1.605305790901184, | |
| "learning_rate": 2.2989658897978082e-05, | |
| "loss": 1.9557, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.7275006965728616, | |
| "grad_norm": 2.3347930908203125, | |
| "learning_rate": 2.2025003858620156e-05, | |
| "loss": 1.9482, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 1.7832265254945667, | |
| "grad_norm": 1.8029910326004028, | |
| "learning_rate": 2.1060348819262234e-05, | |
| "loss": 1.9476, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.7832265254945667, | |
| "eval_loss": 2.0546956062316895, | |
| "eval_rouge1": 0.31858576030641184, | |
| "eval_rouge2": 0.09976539827685843, | |
| "eval_rougeL": 0.1858053683814233, | |
| "eval_rougeLsum": 0.2971687535876159, | |
| "eval_runtime": 6388.7784, | |
| "eval_samples_per_second": 2.092, | |
| "eval_steps_per_second": 0.131, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.8389523544162718, | |
| "grad_norm": 3.6298651695251465, | |
| "learning_rate": 2.0095693779904308e-05, | |
| "loss": 1.9399, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 1.894678183337977, | |
| "grad_norm": 1.8237981796264648, | |
| "learning_rate": 1.9131038740546382e-05, | |
| "loss": 1.9255, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.9504040122596824, | |
| "grad_norm": 1.8469113111495972, | |
| "learning_rate": 1.8166383701188456e-05, | |
| "loss": 1.9222, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 2.0061298411813877, | |
| "grad_norm": 2.2777647972106934, | |
| "learning_rate": 1.720172866183053e-05, | |
| "loss": 1.8964, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.0061298411813877, | |
| "eval_loss": 2.0509514808654785, | |
| "eval_rouge1": 0.3031551742963028, | |
| "eval_rouge2": 0.09186053471142167, | |
| "eval_rougeL": 0.17780386601249287, | |
| "eval_rougeLsum": 0.28264405636299283, | |
| "eval_runtime": 6380.9224, | |
| "eval_samples_per_second": 2.095, | |
| "eval_steps_per_second": 0.131, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.0618556701030926, | |
| "grad_norm": 21.444568634033203, | |
| "learning_rate": 1.6237073622472604e-05, | |
| "loss": 1.6354, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 2.117581499024798, | |
| "grad_norm": 2.026742935180664, | |
| "learning_rate": 1.5272418583114678e-05, | |
| "loss": 1.6349, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 2.173307327946503, | |
| "grad_norm": 2.846383571624756, | |
| "learning_rate": 1.4307763543756752e-05, | |
| "loss": 1.6404, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 2.2290331568682085, | |
| "grad_norm": 2.181227684020996, | |
| "learning_rate": 1.3343108504398828e-05, | |
| "loss": 1.6389, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.2290331568682085, | |
| "eval_loss": 2.048452615737915, | |
| "eval_rouge1": 0.31852665424090676, | |
| "eval_rouge2": 0.0997116570016274, | |
| "eval_rougeL": 0.1856553825562861, | |
| "eval_rougeLsum": 0.29617151453401813, | |
| "eval_runtime": 6398.3205, | |
| "eval_samples_per_second": 2.089, | |
| "eval_steps_per_second": 0.131, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.2847589857899138, | |
| "grad_norm": 2.3432796001434326, | |
| "learning_rate": 1.2378453465040902e-05, | |
| "loss": 1.63, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 2.340484814711619, | |
| "grad_norm": 3.7577052116394043, | |
| "learning_rate": 1.1413798425682977e-05, | |
| "loss": 1.6345, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 2.396210643633324, | |
| "grad_norm": 2.6469638347625732, | |
| "learning_rate": 1.0449143386325052e-05, | |
| "loss": 1.6257, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 2.4519364725550292, | |
| "grad_norm": 1.9295642375946045, | |
| "learning_rate": 9.484488346967126e-06, | |
| "loss": 1.6251, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 2.4519364725550292, | |
| "eval_loss": 2.0327651500701904, | |
| "eval_rouge1": 0.2991586762996594, | |
| "eval_rouge2": 0.0889418449780405, | |
| "eval_rougeL": 0.1742592763849186, | |
| "eval_rougeLsum": 0.2782439206965389, | |
| "eval_runtime": 6392.6136, | |
| "eval_samples_per_second": 2.091, | |
| "eval_steps_per_second": 0.131, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 2.5076623014767345, | |
| "grad_norm": 3.863924264907837, | |
| "learning_rate": 8.5198333076092e-06, | |
| "loss": 1.6185, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 2.56338813039844, | |
| "grad_norm": 13.475985527038574, | |
| "learning_rate": 7.5551782682512745e-06, | |
| "loss": 1.6168, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 2.6191139593201447, | |
| "grad_norm": 20.322145462036133, | |
| "learning_rate": 6.5905232288933485e-06, | |
| "loss": 1.6173, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 2.67483978824185, | |
| "grad_norm": 2.083725929260254, | |
| "learning_rate": 5.6258681895354226e-06, | |
| "loss": 1.6119, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 2.67483978824185, | |
| "eval_loss": 2.013817310333252, | |
| "eval_rouge1": 0.30453205743071354, | |
| "eval_rouge2": 0.0939110545351104, | |
| "eval_rougeL": 0.17781639483308936, | |
| "eval_rougeLsum": 0.28319401801413485, | |
| "eval_runtime": 6384.3929, | |
| "eval_samples_per_second": 2.094, | |
| "eval_steps_per_second": 0.131, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 2.7305656171635553, | |
| "grad_norm": 2.3929226398468018, | |
| "learning_rate": 4.661213150177497e-06, | |
| "loss": 1.6204, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 2.7862914460852606, | |
| "grad_norm": 2.083616256713867, | |
| "learning_rate": 3.6965581108195706e-06, | |
| "loss": 1.6091, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 2.842017275006966, | |
| "grad_norm": 1.816172480583191, | |
| "learning_rate": 2.7319030714616455e-06, | |
| "loss": 1.6104, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 2.897743103928671, | |
| "grad_norm": 7.610255241394043, | |
| "learning_rate": 1.7672480321037198e-06, | |
| "loss": 1.6093, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 2.897743103928671, | |
| "eval_loss": 2.0071072578430176, | |
| "eval_rouge1": 0.30614503423795814, | |
| "eval_rouge2": 0.09393803869603295, | |
| "eval_rougeL": 0.17798792217037074, | |
| "eval_rougeLsum": 0.28484623142325816, | |
| "eval_runtime": 6392.1852, | |
| "eval_samples_per_second": 2.091, | |
| "eval_steps_per_second": 0.131, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 2.953468932850376, | |
| "grad_norm": 5.285951137542725, | |
| "learning_rate": 8.025929927457941e-07, | |
| "loss": 1.6078, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 2.999832822513235, | |
| "step": 26916, | |
| "total_flos": 4.3736503104621773e+18, | |
| "train_loss": 2.05411821950118, | |
| "train_runtime": 108823.6651, | |
| "train_samples_per_second": 7.915, | |
| "train_steps_per_second": 0.247 | |
| } | |
| ], | |
| "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 | |
| } | |