| { | |
| "best_metric": 0.5404770374298096, | |
| "best_model_checkpoint": "/scratch/skscla001/speech/results/whisper-medium-swagen-female-model/checkpoint-1200", | |
| "epoch": 4.225616921269095, | |
| "eval_steps": 200, | |
| "global_step": 1800, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05875440658049354, | |
| "grad_norm": 73.04696655273438, | |
| "learning_rate": 4.2000000000000006e-07, | |
| "loss": 5.3434, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.11750881316098707, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.000000000000001e-07, | |
| "loss": 4.1395, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1762632197414806, | |
| "grad_norm": 34.922523498535156, | |
| "learning_rate": 1.4000000000000001e-06, | |
| "loss": 3.1888, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.23501762632197415, | |
| "grad_norm": 34.749629974365234, | |
| "learning_rate": 1.9000000000000002e-06, | |
| "loss": 2.3154, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2937720329024677, | |
| "grad_norm": 33.44667053222656, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 1.691, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.3525264394829612, | |
| "grad_norm": 38.09309768676758, | |
| "learning_rate": 2.9e-06, | |
| "loss": 1.5853, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.4112808460634548, | |
| "grad_norm": 31.004915237426758, | |
| "learning_rate": 3.4000000000000005e-06, | |
| "loss": 1.522, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.4700352526439483, | |
| "grad_norm": 31.333637237548828, | |
| "learning_rate": 3.900000000000001e-06, | |
| "loss": 1.4291, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4700352526439483, | |
| "eval_loss": 0.8216601610183716, | |
| "eval_runtime": 224.1212, | |
| "eval_samples_per_second": 2.579, | |
| "eval_steps_per_second": 0.647, | |
| "eval_wer": 0.49672320740169623, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5287896592244419, | |
| "grad_norm": 34.124698638916016, | |
| "learning_rate": 4.4e-06, | |
| "loss": 1.2954, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.5875440658049353, | |
| "grad_norm": 38.50987243652344, | |
| "learning_rate": 4.9000000000000005e-06, | |
| "loss": 1.2541, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.6462984723854289, | |
| "grad_norm": 38.04581069946289, | |
| "learning_rate": 5.400000000000001e-06, | |
| "loss": 1.1264, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.7050528789659224, | |
| "grad_norm": 28.410049438476562, | |
| "learning_rate": 5.9e-06, | |
| "loss": 1.1311, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.763807285546416, | |
| "grad_norm": 30.25857925415039, | |
| "learning_rate": 6.4000000000000006e-06, | |
| "loss": 1.0838, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.8225616921269095, | |
| "grad_norm": 31.186445236206055, | |
| "learning_rate": 6.9e-06, | |
| "loss": 1.1503, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.881316098707403, | |
| "grad_norm": 24.111587524414062, | |
| "learning_rate": 7.4e-06, | |
| "loss": 1.0892, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.9400705052878966, | |
| "grad_norm": 22.854475021362305, | |
| "learning_rate": 7.9e-06, | |
| "loss": 0.9176, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.9400705052878966, | |
| "eval_loss": 0.634784460067749, | |
| "eval_runtime": 226.4974, | |
| "eval_samples_per_second": 2.552, | |
| "eval_steps_per_second": 0.64, | |
| "eval_wer": 0.42656129529683884, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.9988249118683902, | |
| "grad_norm": 36.47032928466797, | |
| "learning_rate": 8.400000000000001e-06, | |
| "loss": 1.0234, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.0564042303172738, | |
| "grad_norm": 21.01776123046875, | |
| "learning_rate": 8.900000000000001e-06, | |
| "loss": 0.5964, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.1151586368977673, | |
| "grad_norm": 17.60657501220703, | |
| "learning_rate": 9.4e-06, | |
| "loss": 0.7129, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.1739130434782608, | |
| "grad_norm": 19.744020462036133, | |
| "learning_rate": 9.9e-06, | |
| "loss": 0.5569, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.2326674500587544, | |
| "grad_norm": 15.34373664855957, | |
| "learning_rate": 9.955555555555556e-06, | |
| "loss": 0.5868, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.291421856639248, | |
| "grad_norm": 26.74762535095215, | |
| "learning_rate": 9.9e-06, | |
| "loss": 0.6199, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.3501762632197414, | |
| "grad_norm": 13.160558700561523, | |
| "learning_rate": 9.844444444444446e-06, | |
| "loss": 0.557, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.408930669800235, | |
| "grad_norm": 14.749505043029785, | |
| "learning_rate": 9.78888888888889e-06, | |
| "loss": 0.5492, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.408930669800235, | |
| "eval_loss": 0.5867528319358826, | |
| "eval_runtime": 230.3509, | |
| "eval_samples_per_second": 2.509, | |
| "eval_steps_per_second": 0.629, | |
| "eval_wer": 0.40940632228218965, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.4676850763807285, | |
| "grad_norm": 16.107746124267578, | |
| "learning_rate": 9.733333333333334e-06, | |
| "loss": 0.5442, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.526439482961222, | |
| "grad_norm": 29.692367553710938, | |
| "learning_rate": 9.677777777777778e-06, | |
| "loss": 0.6473, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.5851938895417157, | |
| "grad_norm": 18.0280704498291, | |
| "learning_rate": 9.622222222222222e-06, | |
| "loss": 0.5739, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.6439482961222092, | |
| "grad_norm": 27.483768463134766, | |
| "learning_rate": 9.566666666666668e-06, | |
| "loss": 0.6705, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.7027027027027026, | |
| "grad_norm": 20.081012725830078, | |
| "learning_rate": 9.511111111111112e-06, | |
| "loss": 0.6338, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.7614571092831963, | |
| "grad_norm": 22.985151290893555, | |
| "learning_rate": 9.455555555555557e-06, | |
| "loss": 0.6409, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.8202115158636898, | |
| "grad_norm": 19.641082763671875, | |
| "learning_rate": 9.4e-06, | |
| "loss": 0.6777, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.8789659224441833, | |
| "grad_norm": 27.92637825012207, | |
| "learning_rate": 9.344444444444446e-06, | |
| "loss": 0.6243, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.8789659224441833, | |
| "eval_loss": 0.5535122752189636, | |
| "eval_runtime": 218.1215, | |
| "eval_samples_per_second": 2.65, | |
| "eval_steps_per_second": 0.665, | |
| "eval_wer": 0.3274865073245952, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.937720329024677, | |
| "grad_norm": 24.763235092163086, | |
| "learning_rate": 9.28888888888889e-06, | |
| "loss": 0.5296, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.9964747356051704, | |
| "grad_norm": 22.559097290039062, | |
| "learning_rate": 9.233333333333334e-06, | |
| "loss": 0.6109, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.054054054054054, | |
| "grad_norm": 9.163862228393555, | |
| "learning_rate": 9.17777777777778e-06, | |
| "loss": 0.2217, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 2.1128084606345476, | |
| "grad_norm": 14.249441146850586, | |
| "learning_rate": 9.122222222222223e-06, | |
| "loss": 0.2233, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.1715628672150413, | |
| "grad_norm": 12.58755111694336, | |
| "learning_rate": 9.066666666666667e-06, | |
| "loss": 0.2103, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 2.2303172737955346, | |
| "grad_norm": 12.163936614990234, | |
| "learning_rate": 9.011111111111111e-06, | |
| "loss": 0.2184, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.2890716803760283, | |
| "grad_norm": 24.493022918701172, | |
| "learning_rate": 8.955555555555555e-06, | |
| "loss": 0.2056, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 2.3478260869565215, | |
| "grad_norm": 16.613021850585938, | |
| "learning_rate": 8.900000000000001e-06, | |
| "loss": 0.2196, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.3478260869565215, | |
| "eval_loss": 0.5643105506896973, | |
| "eval_runtime": 222.0189, | |
| "eval_samples_per_second": 2.603, | |
| "eval_steps_per_second": 0.653, | |
| "eval_wer": 0.35774865073245954, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.406580493537015, | |
| "grad_norm": 9.612153053283691, | |
| "learning_rate": 8.844444444444445e-06, | |
| "loss": 0.2273, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 2.465334900117509, | |
| "grad_norm": 8.410961151123047, | |
| "learning_rate": 8.788888888888891e-06, | |
| "loss": 0.2034, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.524089306698002, | |
| "grad_norm": 21.08755874633789, | |
| "learning_rate": 8.733333333333333e-06, | |
| "loss": 0.2239, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 2.582843713278496, | |
| "grad_norm": 14.133148193359375, | |
| "learning_rate": 8.677777777777779e-06, | |
| "loss": 0.2086, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.6415981198589895, | |
| "grad_norm": 6.469407558441162, | |
| "learning_rate": 8.622222222222223e-06, | |
| "loss": 0.2349, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 2.7003525264394828, | |
| "grad_norm": 12.72140121459961, | |
| "learning_rate": 8.566666666666667e-06, | |
| "loss": 0.2101, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.7591069330199764, | |
| "grad_norm": 7.142697334289551, | |
| "learning_rate": 8.511111111111113e-06, | |
| "loss": 0.2615, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 2.81786133960047, | |
| "grad_norm": 11.419723510742188, | |
| "learning_rate": 8.455555555555555e-06, | |
| "loss": 0.2211, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.81786133960047, | |
| "eval_loss": 0.5404770374298096, | |
| "eval_runtime": 217.6495, | |
| "eval_samples_per_second": 2.656, | |
| "eval_steps_per_second": 0.666, | |
| "eval_wer": 0.33982266769468006, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.8766157461809634, | |
| "grad_norm": 8.013970375061035, | |
| "learning_rate": 8.400000000000001e-06, | |
| "loss": 0.2023, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 2.935370152761457, | |
| "grad_norm": 11.708134651184082, | |
| "learning_rate": 8.344444444444445e-06, | |
| "loss": 0.2406, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.9941245593419508, | |
| "grad_norm": 13.683923721313477, | |
| "learning_rate": 8.288888888888889e-06, | |
| "loss": 0.2345, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 3.0517038777908345, | |
| "grad_norm": 4.092015743255615, | |
| "learning_rate": 8.233333333333335e-06, | |
| "loss": 0.0736, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 3.1104582843713278, | |
| "grad_norm": 7.464056015014648, | |
| "learning_rate": 8.177777777777779e-06, | |
| "loss": 0.0939, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 3.1692126909518215, | |
| "grad_norm": 6.3624677658081055, | |
| "learning_rate": 8.122222222222223e-06, | |
| "loss": 0.0829, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 3.227967097532315, | |
| "grad_norm": 9.64356803894043, | |
| "learning_rate": 8.066666666666667e-06, | |
| "loss": 0.0939, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 3.2867215041128084, | |
| "grad_norm": 9.63764476776123, | |
| "learning_rate": 8.011111111111113e-06, | |
| "loss": 0.0999, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 3.2867215041128084, | |
| "eval_loss": 0.5825645923614502, | |
| "eval_runtime": 218.2885, | |
| "eval_samples_per_second": 2.648, | |
| "eval_steps_per_second": 0.664, | |
| "eval_wer": 0.3282575173477255, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 3.345475910693302, | |
| "grad_norm": 6.401581764221191, | |
| "learning_rate": 7.955555555555557e-06, | |
| "loss": 0.0829, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 3.4042303172737958, | |
| "grad_norm": 6.341639041900635, | |
| "learning_rate": 7.9e-06, | |
| "loss": 0.0914, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 3.462984723854289, | |
| "grad_norm": 5.883626937866211, | |
| "learning_rate": 7.844444444444446e-06, | |
| "loss": 0.1057, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 3.5217391304347827, | |
| "grad_norm": 10.137381553649902, | |
| "learning_rate": 7.788888888888889e-06, | |
| "loss": 0.0948, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.5804935370152764, | |
| "grad_norm": 3.2198357582092285, | |
| "learning_rate": 7.733333333333334e-06, | |
| "loss": 0.0916, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 3.6392479435957696, | |
| "grad_norm": 4.958520412445068, | |
| "learning_rate": 7.677777777777778e-06, | |
| "loss": 0.0847, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 3.6980023501762633, | |
| "grad_norm": 7.396123886108398, | |
| "learning_rate": 7.622222222222223e-06, | |
| "loss": 0.1091, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 3.756756756756757, | |
| "grad_norm": 8.138400077819824, | |
| "learning_rate": 7.566666666666667e-06, | |
| "loss": 0.1111, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.756756756756757, | |
| "eval_loss": 0.5536873936653137, | |
| "eval_runtime": 217.469, | |
| "eval_samples_per_second": 2.658, | |
| "eval_steps_per_second": 0.667, | |
| "eval_wer": 0.3276792598303778, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.8155111633372503, | |
| "grad_norm": 7.745422840118408, | |
| "learning_rate": 7.511111111111111e-06, | |
| "loss": 0.1031, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 3.874265569917744, | |
| "grad_norm": 7.68823766708374, | |
| "learning_rate": 7.455555555555556e-06, | |
| "loss": 0.0919, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 3.933019976498237, | |
| "grad_norm": 4.057219505310059, | |
| "learning_rate": 7.4e-06, | |
| "loss": 0.1069, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 3.991774383078731, | |
| "grad_norm": 7.178942680358887, | |
| "learning_rate": 7.344444444444445e-06, | |
| "loss": 0.0951, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 4.049353701527615, | |
| "grad_norm": 5.237444877624512, | |
| "learning_rate": 7.28888888888889e-06, | |
| "loss": 0.0495, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 4.108108108108108, | |
| "grad_norm": 7.6568684577941895, | |
| "learning_rate": 7.233333333333334e-06, | |
| "loss": 0.0546, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 4.166862514688602, | |
| "grad_norm": 5.4397149085998535, | |
| "learning_rate": 7.177777777777778e-06, | |
| "loss": 0.0521, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 4.225616921269095, | |
| "grad_norm": 7.218142032623291, | |
| "learning_rate": 7.122222222222222e-06, | |
| "loss": 0.0423, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 4.225616921269095, | |
| "eval_loss": 0.6012063026428223, | |
| "eval_runtime": 217.9147, | |
| "eval_samples_per_second": 2.652, | |
| "eval_steps_per_second": 0.665, | |
| "eval_wer": 0.31881264456437935, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 4.225616921269095, | |
| "step": 1800, | |
| "total_flos": 1.468038098976768e+19, | |
| "train_loss": 0.6555014891094632, | |
| "train_runtime": 4980.8387, | |
| "train_samples_per_second": 8.031, | |
| "train_steps_per_second": 1.004 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 5000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 12, | |
| "save_steps": 200, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 3, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 3 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.468038098976768e+19, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |