| { | |
| "best_metric": 0.6367625594139099, | |
| "best_model_checkpoint": "/scratch/skscla001/speech/results/whisper-medium-bsbigcgen-female-model/checkpoint-600", | |
| "epoch": 3.6933744221879814, | |
| "eval_steps": 200, | |
| "global_step": 1200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.07704160246533127, | |
| "grad_norm": 112.25947570800781, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "loss": 11.5295, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.15408320493066255, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8.8e-07, | |
| "loss": 9.8901, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.23112480739599384, | |
| "grad_norm": 84.68060302734375, | |
| "learning_rate": 1.3800000000000001e-06, | |
| "loss": 7.8133, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.3081664098613251, | |
| "grad_norm": 94.15545654296875, | |
| "learning_rate": 1.8800000000000002e-06, | |
| "loss": 6.4974, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3852080123266564, | |
| "grad_norm": 72.34725189208984, | |
| "learning_rate": 2.38e-06, | |
| "loss": 4.9676, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.4622496147919877, | |
| "grad_norm": 67.70198059082031, | |
| "learning_rate": 2.88e-06, | |
| "loss": 4.2556, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.539291217257319, | |
| "grad_norm": 80.96290588378906, | |
| "learning_rate": 3.3800000000000007e-06, | |
| "loss": 4.3249, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.6163328197226502, | |
| "grad_norm": 67.85588073730469, | |
| "learning_rate": 3.88e-06, | |
| "loss": 3.47, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6163328197226502, | |
| "eval_loss": 0.927435576915741, | |
| "eval_runtime": 249.4489, | |
| "eval_samples_per_second": 1.728, | |
| "eval_steps_per_second": 0.866, | |
| "eval_wer": 0.6674612634088201, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6933744221879815, | |
| "grad_norm": 60.49135971069336, | |
| "learning_rate": 4.38e-06, | |
| "loss": 3.7208, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.7704160246533128, | |
| "grad_norm": 84.06492614746094, | |
| "learning_rate": 4.880000000000001e-06, | |
| "loss": 3.6644, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.847457627118644, | |
| "grad_norm": 66.79823303222656, | |
| "learning_rate": 5.380000000000001e-06, | |
| "loss": 3.1538, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.9244992295839753, | |
| "grad_norm": 66.50402069091797, | |
| "learning_rate": 5.8800000000000005e-06, | |
| "loss": 3.0771, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 68.71908569335938, | |
| "learning_rate": 6.380000000000001e-06, | |
| "loss": 3.2809, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.0770416024653313, | |
| "grad_norm": 51.007381439208984, | |
| "learning_rate": 6.88e-06, | |
| "loss": 2.4943, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.1540832049306626, | |
| "grad_norm": 59.46632385253906, | |
| "learning_rate": 7.3800000000000005e-06, | |
| "loss": 2.2635, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.231124807395994, | |
| "grad_norm": 61.94889831542969, | |
| "learning_rate": 7.88e-06, | |
| "loss": 2.569, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.231124807395994, | |
| "eval_loss": 0.7236458659172058, | |
| "eval_runtime": 242.3506, | |
| "eval_samples_per_second": 1.778, | |
| "eval_steps_per_second": 0.891, | |
| "eval_wer": 0.5511323003575685, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.308166409861325, | |
| "grad_norm": 54.65010070800781, | |
| "learning_rate": 8.380000000000001e-06, | |
| "loss": 2.4459, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.3852080123266564, | |
| "grad_norm": 69.15284729003906, | |
| "learning_rate": 8.880000000000001e-06, | |
| "loss": 2.1676, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.4622496147919877, | |
| "grad_norm": 40.50135040283203, | |
| "learning_rate": 9.38e-06, | |
| "loss": 2.1835, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.539291217257319, | |
| "grad_norm": 47.11216735839844, | |
| "learning_rate": 9.88e-06, | |
| "loss": 2.7759, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.61633281972265, | |
| "grad_norm": 41.89402389526367, | |
| "learning_rate": 9.979392624728852e-06, | |
| "loss": 2.5969, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.6933744221879814, | |
| "grad_norm": 54.26713562011719, | |
| "learning_rate": 9.952277657266813e-06, | |
| "loss": 2.0526, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.7704160246533127, | |
| "grad_norm": 48.94434356689453, | |
| "learning_rate": 9.925162689804773e-06, | |
| "loss": 2.068, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.847457627118644, | |
| "grad_norm": 46.54755783081055, | |
| "learning_rate": 9.898047722342734e-06, | |
| "loss": 2.5305, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.847457627118644, | |
| "eval_loss": 0.6367625594139099, | |
| "eval_runtime": 236.1921, | |
| "eval_samples_per_second": 1.825, | |
| "eval_steps_per_second": 0.915, | |
| "eval_wer": 0.47723480333730633, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.9244992295839753, | |
| "grad_norm": 45.87973403930664, | |
| "learning_rate": 9.870932754880696e-06, | |
| "loss": 2.4065, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 28.206829071044922, | |
| "learning_rate": 9.843817787418655e-06, | |
| "loss": 2.0807, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.0770416024653313, | |
| "grad_norm": 29.09613800048828, | |
| "learning_rate": 9.816702819956617e-06, | |
| "loss": 1.3675, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 2.1540832049306626, | |
| "grad_norm": 28.714885711669922, | |
| "learning_rate": 9.789587852494578e-06, | |
| "loss": 1.2501, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.231124807395994, | |
| "grad_norm": 54.41804122924805, | |
| "learning_rate": 9.76247288503254e-06, | |
| "loss": 1.2393, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 2.3081664098613253, | |
| "grad_norm": 31.56501007080078, | |
| "learning_rate": 9.7353579175705e-06, | |
| "loss": 1.2768, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.3852080123266566, | |
| "grad_norm": 73.83405303955078, | |
| "learning_rate": 9.70824295010846e-06, | |
| "loss": 1.2799, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 2.462249614791988, | |
| "grad_norm": 42.22137451171875, | |
| "learning_rate": 9.68112798264642e-06, | |
| "loss": 1.4563, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.462249614791988, | |
| "eval_loss": 0.651836097240448, | |
| "eval_runtime": 241.7933, | |
| "eval_samples_per_second": 1.783, | |
| "eval_steps_per_second": 0.893, | |
| "eval_wer": 0.4936829558998808, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.539291217257319, | |
| "grad_norm": 58.9369010925293, | |
| "learning_rate": 9.654013015184382e-06, | |
| "loss": 1.4182, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 2.61633281972265, | |
| "grad_norm": 33.295654296875, | |
| "learning_rate": 9.626898047722343e-06, | |
| "loss": 1.5956, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.6933744221879814, | |
| "grad_norm": 31.526493072509766, | |
| "learning_rate": 9.599783080260305e-06, | |
| "loss": 1.4829, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 2.7704160246533127, | |
| "grad_norm": 38.276947021484375, | |
| "learning_rate": 9.572668112798266e-06, | |
| "loss": 1.4473, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.847457627118644, | |
| "grad_norm": 35.50004959106445, | |
| "learning_rate": 9.545553145336226e-06, | |
| "loss": 1.417, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 2.9244992295839753, | |
| "grad_norm": 29.786792755126953, | |
| "learning_rate": 9.518438177874187e-06, | |
| "loss": 1.2574, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 12.207977294921875, | |
| "learning_rate": 9.491323210412147e-06, | |
| "loss": 1.2935, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 3.0770416024653313, | |
| "grad_norm": 27.92531394958496, | |
| "learning_rate": 9.464208242950108e-06, | |
| "loss": 0.6413, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.0770416024653313, | |
| "eval_loss": 0.6891891956329346, | |
| "eval_runtime": 240.7265, | |
| "eval_samples_per_second": 1.79, | |
| "eval_steps_per_second": 0.897, | |
| "eval_wer": 0.47818831942789036, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.1540832049306626, | |
| "grad_norm": 30.59457778930664, | |
| "learning_rate": 9.43709327548807e-06, | |
| "loss": 0.764, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 3.231124807395994, | |
| "grad_norm": 23.21906852722168, | |
| "learning_rate": 9.409978308026031e-06, | |
| "loss": 0.6408, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 3.3081664098613253, | |
| "grad_norm": 28.22869110107422, | |
| "learning_rate": 9.382863340563993e-06, | |
| "loss": 0.6615, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 3.3852080123266566, | |
| "grad_norm": 32.756710052490234, | |
| "learning_rate": 9.355748373101952e-06, | |
| "loss": 0.7282, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.462249614791988, | |
| "grad_norm": 27.851184844970703, | |
| "learning_rate": 9.328633405639914e-06, | |
| "loss": 0.8159, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 3.539291217257319, | |
| "grad_norm": 39.99174118041992, | |
| "learning_rate": 9.301518438177875e-06, | |
| "loss": 1.218, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 3.61633281972265, | |
| "grad_norm": 28.462491989135742, | |
| "learning_rate": 9.274403470715837e-06, | |
| "loss": 0.7496, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 3.6933744221879814, | |
| "grad_norm": 30.86623191833496, | |
| "learning_rate": 9.247288503253798e-06, | |
| "loss": 0.581, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.6933744221879814, | |
| "eval_loss": 0.6935437321662903, | |
| "eval_runtime": 235.9787, | |
| "eval_samples_per_second": 1.826, | |
| "eval_steps_per_second": 0.915, | |
| "eval_wer": 0.47508939213349227, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.6933744221879814, | |
| "step": 1200, | |
| "total_flos": 9.7824980385792e+18, | |
| "train_loss": 2.601292386849721, | |
| "train_runtime": 3847.301, | |
| "train_samples_per_second": 20.235, | |
| "train_steps_per_second": 2.526 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 9720, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 30, | |
| "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": 9.7824980385792e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |