| { | |
| "best_global_step": 600, | |
| "best_metric": 0.6628760695457458, | |
| "best_model_checkpoint": "/scratch/skscla001/experiments/datasets/results/whisper-medium-bigcgen-male-5hrs-62/checkpoint-600", | |
| "epoch": 4.3478260869565215, | |
| "eval_steps": 200, | |
| "global_step": 1400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.07763975155279502, | |
| "grad_norm": 33.341270446777344, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "loss": 3.3362, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.15527950310559005, | |
| "grad_norm": 26.828624725341797, | |
| "learning_rate": 9.000000000000001e-07, | |
| "loss": 2.6418, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.2329192546583851, | |
| "grad_norm": 25.673059463500977, | |
| "learning_rate": 1.4000000000000001e-06, | |
| "loss": 2.017, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.3105590062111801, | |
| "grad_norm": 17.314512252807617, | |
| "learning_rate": 1.9000000000000002e-06, | |
| "loss": 1.4498, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.38819875776397517, | |
| "grad_norm": 17.275650024414062, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 1.2023, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.4658385093167702, | |
| "grad_norm": 17.702608108520508, | |
| "learning_rate": 2.9e-06, | |
| "loss": 1.0835, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5434782608695652, | |
| "grad_norm": 15.095105171203613, | |
| "learning_rate": 3.4000000000000005e-06, | |
| "loss": 1.1593, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.6211180124223602, | |
| "grad_norm": 16.568078994750977, | |
| "learning_rate": 3.900000000000001e-06, | |
| "loss": 0.956, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6211180124223602, | |
| "eval_loss": 0.8870997428894043, | |
| "eval_runtime": 276.5554, | |
| "eval_samples_per_second": 1.595, | |
| "eval_steps_per_second": 0.799, | |
| "eval_wer": 0.6443856696321231, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6987577639751553, | |
| "grad_norm": 17.126779556274414, | |
| "learning_rate": 4.4e-06, | |
| "loss": 1.0133, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.7763975155279503, | |
| "grad_norm": 15.565762519836426, | |
| "learning_rate": 4.9000000000000005e-06, | |
| "loss": 1.0108, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8540372670807453, | |
| "grad_norm": 19.874563217163086, | |
| "learning_rate": 5.400000000000001e-06, | |
| "loss": 0.792, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.9316770186335404, | |
| "grad_norm": 13.371776580810547, | |
| "learning_rate": 5.9e-06, | |
| "loss": 0.7788, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.0093167701863355, | |
| "grad_norm": 13.305375099182129, | |
| "learning_rate": 6.4000000000000006e-06, | |
| "loss": 0.8068, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.0869565217391304, | |
| "grad_norm": 16.924894332885742, | |
| "learning_rate": 6.9e-06, | |
| "loss": 0.5933, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.1645962732919255, | |
| "grad_norm": 13.33420467376709, | |
| "learning_rate": 7.4e-06, | |
| "loss": 0.7016, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.2422360248447206, | |
| "grad_norm": 15.580875396728516, | |
| "learning_rate": 7.9e-06, | |
| "loss": 0.6779, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.2422360248447206, | |
| "eval_loss": 0.7327317595481873, | |
| "eval_runtime": 265.8594, | |
| "eval_samples_per_second": 1.659, | |
| "eval_steps_per_second": 0.831, | |
| "eval_wer": 0.5585477278191873, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.3198757763975155, | |
| "grad_norm": 14.909109115600586, | |
| "learning_rate": 8.400000000000001e-06, | |
| "loss": 0.6421, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.3975155279503104, | |
| "grad_norm": 15.768739700317383, | |
| "learning_rate": 8.900000000000001e-06, | |
| "loss": 0.6492, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.4751552795031055, | |
| "grad_norm": 10.104466438293457, | |
| "learning_rate": 9.4e-06, | |
| "loss": 0.6439, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.5527950310559007, | |
| "grad_norm": 12.118277549743652, | |
| "learning_rate": 9.9e-06, | |
| "loss": 0.6893, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.6304347826086958, | |
| "grad_norm": 11.13495922088623, | |
| "learning_rate": 9.955555555555556e-06, | |
| "loss": 0.6307, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.7080745341614907, | |
| "grad_norm": 12.645089149475098, | |
| "learning_rate": 9.9e-06, | |
| "loss": 0.603, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.7857142857142856, | |
| "grad_norm": 17.46633529663086, | |
| "learning_rate": 9.844444444444446e-06, | |
| "loss": 0.657, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.8633540372670807, | |
| "grad_norm": 8.319685935974121, | |
| "learning_rate": 9.78888888888889e-06, | |
| "loss": 0.5879, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.8633540372670807, | |
| "eval_loss": 0.6628760695457458, | |
| "eval_runtime": 270.2431, | |
| "eval_samples_per_second": 1.632, | |
| "eval_steps_per_second": 0.818, | |
| "eval_wer": 0.516229862947824, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.9409937888198758, | |
| "grad_norm": 11.236241340637207, | |
| "learning_rate": 9.733333333333334e-06, | |
| "loss": 0.6359, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 2.018633540372671, | |
| "grad_norm": 13.337384223937988, | |
| "learning_rate": 9.677777777777778e-06, | |
| "loss": 0.5992, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.0962732919254656, | |
| "grad_norm": 8.325053215026855, | |
| "learning_rate": 9.622222222222222e-06, | |
| "loss": 0.3897, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 2.1739130434782608, | |
| "grad_norm": 12.123583793640137, | |
| "learning_rate": 9.566666666666668e-06, | |
| "loss": 0.4313, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.251552795031056, | |
| "grad_norm": 13.106657028198242, | |
| "learning_rate": 9.511111111111112e-06, | |
| "loss": 0.3598, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 2.329192546583851, | |
| "grad_norm": 8.855884552001953, | |
| "learning_rate": 9.455555555555557e-06, | |
| "loss": 0.4473, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.406832298136646, | |
| "grad_norm": 8.3619966506958, | |
| "learning_rate": 9.4e-06, | |
| "loss": 0.3956, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 2.4844720496894412, | |
| "grad_norm": 9.356673240661621, | |
| "learning_rate": 9.344444444444446e-06, | |
| "loss": 0.4268, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.4844720496894412, | |
| "eval_loss": 0.6734069585800171, | |
| "eval_runtime": 275.7178, | |
| "eval_samples_per_second": 1.599, | |
| "eval_steps_per_second": 0.802, | |
| "eval_wer": 0.5006011060351045, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.562111801242236, | |
| "grad_norm": 6.666379928588867, | |
| "learning_rate": 9.28888888888889e-06, | |
| "loss": 0.389, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 2.639751552795031, | |
| "grad_norm": 11.296941757202148, | |
| "learning_rate": 9.233333333333334e-06, | |
| "loss": 0.3558, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.717391304347826, | |
| "grad_norm": 14.280022621154785, | |
| "learning_rate": 9.17777777777778e-06, | |
| "loss": 0.4083, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 2.795031055900621, | |
| "grad_norm": 11.778422355651855, | |
| "learning_rate": 9.122222222222223e-06, | |
| "loss": 0.4209, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.8726708074534164, | |
| "grad_norm": 8.68419075012207, | |
| "learning_rate": 9.066666666666667e-06, | |
| "loss": 0.3953, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 2.950310559006211, | |
| "grad_norm": 15.859999656677246, | |
| "learning_rate": 9.011111111111111e-06, | |
| "loss": 0.3805, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.027950310559006, | |
| "grad_norm": 6.30320930480957, | |
| "learning_rate": 8.955555555555555e-06, | |
| "loss": 0.3059, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 3.1055900621118013, | |
| "grad_norm": 6.796067714691162, | |
| "learning_rate": 8.900000000000001e-06, | |
| "loss": 0.1979, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.1055900621118013, | |
| "eval_loss": 0.7146615982055664, | |
| "eval_runtime": 265.096, | |
| "eval_samples_per_second": 1.664, | |
| "eval_steps_per_second": 0.834, | |
| "eval_wer": 0.4765568646309209, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.1832298136645965, | |
| "grad_norm": 6.765524864196777, | |
| "learning_rate": 8.844444444444445e-06, | |
| "loss": 0.2458, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 3.260869565217391, | |
| "grad_norm": 6.841522693634033, | |
| "learning_rate": 8.788888888888891e-06, | |
| "loss": 0.2144, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 3.3385093167701863, | |
| "grad_norm": 7.854112148284912, | |
| "learning_rate": 8.733333333333333e-06, | |
| "loss": 0.2255, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 3.4161490683229814, | |
| "grad_norm": 6.754302501678467, | |
| "learning_rate": 8.677777777777779e-06, | |
| "loss": 0.2295, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.4937888198757765, | |
| "grad_norm": 10.24082088470459, | |
| "learning_rate": 8.622222222222223e-06, | |
| "loss": 0.214, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 3.571428571428571, | |
| "grad_norm": 8.713047981262207, | |
| "learning_rate": 8.566666666666667e-06, | |
| "loss": 0.2529, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 3.6490683229813663, | |
| "grad_norm": 10.281318664550781, | |
| "learning_rate": 8.511111111111113e-06, | |
| "loss": 0.2617, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 3.7267080745341614, | |
| "grad_norm": 9.869316101074219, | |
| "learning_rate": 8.455555555555555e-06, | |
| "loss": 0.2338, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.7267080745341614, | |
| "eval_loss": 0.7034153938293457, | |
| "eval_runtime": 264.3393, | |
| "eval_samples_per_second": 1.668, | |
| "eval_steps_per_second": 0.836, | |
| "eval_wer": 0.48689588843471987, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.8043478260869565, | |
| "grad_norm": 4.6300835609436035, | |
| "learning_rate": 8.400000000000001e-06, | |
| "loss": 0.2299, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 3.8819875776397517, | |
| "grad_norm": 6.275107383728027, | |
| "learning_rate": 8.344444444444445e-06, | |
| "loss": 0.1803, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 3.9596273291925463, | |
| "grad_norm": 7.057643413543701, | |
| "learning_rate": 8.288888888888889e-06, | |
| "loss": 0.2024, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 4.037267080745342, | |
| "grad_norm": 7.044947147369385, | |
| "learning_rate": 8.233333333333335e-06, | |
| "loss": 0.1694, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.114906832298137, | |
| "grad_norm": 5.649941921234131, | |
| "learning_rate": 8.177777777777779e-06, | |
| "loss": 0.1026, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 4.192546583850931, | |
| "grad_norm": 6.779381275177002, | |
| "learning_rate": 8.122222222222223e-06, | |
| "loss": 0.1207, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 4.270186335403727, | |
| "grad_norm": 4.534013271331787, | |
| "learning_rate": 8.066666666666667e-06, | |
| "loss": 0.081, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 4.3478260869565215, | |
| "grad_norm": 10.306833267211914, | |
| "learning_rate": 8.011111111111113e-06, | |
| "loss": 0.1014, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.3478260869565215, | |
| "eval_loss": 0.7553574442863464, | |
| "eval_runtime": 263.3312, | |
| "eval_samples_per_second": 1.675, | |
| "eval_steps_per_second": 0.839, | |
| "eval_wer": 0.4878576580908872, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.3478260869565215, | |
| "step": 1400, | |
| "total_flos": 1.1430774964224e+19, | |
| "train_loss": 0.6237195907320295, | |
| "train_runtime": 4683.4552, | |
| "train_samples_per_second": 8.541, | |
| "train_steps_per_second": 1.068 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 5000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 16, | |
| "save_steps": 200, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 4, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 4 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.1430774964224e+19, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |