|
{ |
|
"best_metric": 42.54887827732228, |
|
"best_model_checkpoint": "./whisper-small-hy-nohup-24-hours/checkpoint-3808", |
|
"epoch": 25.0, |
|
"eval_steps": 500, |
|
"global_step": 5950, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 1.0, |
|
"grad_norm": 8.179900169372559, |
|
"learning_rate": 9.601613716591026e-06, |
|
"loss": 0.4522, |
|
"step": 238 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"eval_loss": 0.25782614946365356, |
|
"eval_runtime": 2089.2119, |
|
"eval_samples_per_second": 1.366, |
|
"eval_steps_per_second": 0.171, |
|
"eval_wer": 55.13109289125146, |
|
"step": 238 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"grad_norm": 10.84644889831543, |
|
"learning_rate": 9.201546478399732e-06, |
|
"loss": 0.1602, |
|
"step": 476 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_loss": 0.22305740416049957, |
|
"eval_runtime": 2112.4722, |
|
"eval_samples_per_second": 1.351, |
|
"eval_steps_per_second": 0.169, |
|
"eval_wer": 48.32417334895035, |
|
"step": 476 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"grad_norm": 10.12069320678711, |
|
"learning_rate": 8.80147924020844e-06, |
|
"loss": 0.0876, |
|
"step": 714 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"eval_loss": 0.22624510526657104, |
|
"eval_runtime": 2115.2735, |
|
"eval_samples_per_second": 1.349, |
|
"eval_steps_per_second": 0.169, |
|
"eval_wer": 45.29236868186323, |
|
"step": 714 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"grad_norm": 9.993618965148926, |
|
"learning_rate": 8.401412002017146e-06, |
|
"loss": 0.047, |
|
"step": 952 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"eval_loss": 0.24928627908229828, |
|
"eval_runtime": 2129.3503, |
|
"eval_samples_per_second": 1.34, |
|
"eval_steps_per_second": 0.168, |
|
"eval_wer": 45.17073610235156, |
|
"step": 952 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"grad_norm": 1.5396642684936523, |
|
"learning_rate": 8.001344763825853e-06, |
|
"loss": 0.0259, |
|
"step": 1190 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"eval_loss": 0.26582393050193787, |
|
"eval_runtime": 2108.6013, |
|
"eval_samples_per_second": 1.353, |
|
"eval_steps_per_second": 0.169, |
|
"eval_wer": 44.13010181097396, |
|
"step": 1190 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"grad_norm": 5.9440460205078125, |
|
"learning_rate": 7.601277525634561e-06, |
|
"loss": 0.0144, |
|
"step": 1428 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"eval_loss": 0.28802964091300964, |
|
"eval_runtime": 2110.1435, |
|
"eval_samples_per_second": 1.352, |
|
"eval_steps_per_second": 0.169, |
|
"eval_wer": 44.05802324533742, |
|
"step": 1428 |
|
}, |
|
{ |
|
"epoch": 7.0, |
|
"grad_norm": 0.6433717608451843, |
|
"learning_rate": 7.201210287443269e-06, |
|
"loss": 0.0091, |
|
"step": 1666 |
|
}, |
|
{ |
|
"epoch": 7.0, |
|
"eval_loss": 0.29965829849243164, |
|
"eval_runtime": 2117.1627, |
|
"eval_samples_per_second": 1.348, |
|
"eval_steps_per_second": 0.169, |
|
"eval_wer": 44.16163618343995, |
|
"step": 1666 |
|
}, |
|
{ |
|
"epoch": 8.0, |
|
"grad_norm": 10.917597770690918, |
|
"learning_rate": 6.801143049251976e-06, |
|
"loss": 0.0052, |
|
"step": 1904 |
|
}, |
|
{ |
|
"epoch": 8.0, |
|
"eval_loss": 0.3222191631793976, |
|
"eval_runtime": 2112.2075, |
|
"eval_samples_per_second": 1.351, |
|
"eval_steps_per_second": 0.169, |
|
"eval_wer": 44.91395621227138, |
|
"step": 1904 |
|
}, |
|
{ |
|
"epoch": 9.0, |
|
"grad_norm": 1.45236337184906, |
|
"learning_rate": 6.401075811060683e-06, |
|
"loss": 0.0035, |
|
"step": 2142 |
|
}, |
|
{ |
|
"epoch": 9.0, |
|
"eval_loss": 0.33322593569755554, |
|
"eval_runtime": 2115.5823, |
|
"eval_samples_per_second": 1.349, |
|
"eval_steps_per_second": 0.169, |
|
"eval_wer": 43.43634561672223, |
|
"step": 2142 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"grad_norm": 0.8597022891044617, |
|
"learning_rate": 6.00100857286939e-06, |
|
"loss": 0.0024, |
|
"step": 2380 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"eval_loss": 0.3342152535915375, |
|
"eval_runtime": 2100.0091, |
|
"eval_samples_per_second": 1.359, |
|
"eval_steps_per_second": 0.17, |
|
"eval_wer": 44.07604288674655, |
|
"step": 2380 |
|
}, |
|
{ |
|
"epoch": 11.0, |
|
"grad_norm": 0.5440101623535156, |
|
"learning_rate": 5.6009413346780976e-06, |
|
"loss": 0.0018, |
|
"step": 2618 |
|
}, |
|
{ |
|
"epoch": 11.0, |
|
"eval_loss": 0.34759968519210815, |
|
"eval_runtime": 2112.6804, |
|
"eval_samples_per_second": 1.35, |
|
"eval_steps_per_second": 0.169, |
|
"eval_wer": 44.47247499774755, |
|
"step": 2618 |
|
}, |
|
{ |
|
"epoch": 12.0, |
|
"grad_norm": 0.05317315086722374, |
|
"learning_rate": 5.200874096486806e-06, |
|
"loss": 0.0009, |
|
"step": 2856 |
|
}, |
|
{ |
|
"epoch": 12.0, |
|
"eval_loss": 0.35896193981170654, |
|
"eval_runtime": 2102.0987, |
|
"eval_samples_per_second": 1.357, |
|
"eval_steps_per_second": 0.17, |
|
"eval_wer": 43.265159023335436, |
|
"step": 2856 |
|
}, |
|
{ |
|
"epoch": 13.0, |
|
"grad_norm": 0.02565290778875351, |
|
"learning_rate": 4.800806858295513e-06, |
|
"loss": 0.0006, |
|
"step": 3094 |
|
}, |
|
{ |
|
"epoch": 13.0, |
|
"eval_loss": 0.36282169818878174, |
|
"eval_runtime": 2101.9828, |
|
"eval_samples_per_second": 1.357, |
|
"eval_steps_per_second": 0.17, |
|
"eval_wer": 42.66600594648167, |
|
"step": 3094 |
|
}, |
|
{ |
|
"epoch": 14.0, |
|
"grad_norm": 0.09528347104787827, |
|
"learning_rate": 4.40073962010422e-06, |
|
"loss": 0.0005, |
|
"step": 3332 |
|
}, |
|
{ |
|
"epoch": 14.0, |
|
"eval_loss": 0.36773309111595154, |
|
"eval_runtime": 2113.7555, |
|
"eval_samples_per_second": 1.35, |
|
"eval_steps_per_second": 0.169, |
|
"eval_wer": 42.81016307775475, |
|
"step": 3332 |
|
}, |
|
{ |
|
"epoch": 15.0, |
|
"grad_norm": 0.025408850982785225, |
|
"learning_rate": 4.000672381912926e-06, |
|
"loss": 0.0004, |
|
"step": 3570 |
|
}, |
|
{ |
|
"epoch": 15.0, |
|
"eval_loss": 0.36659348011016846, |
|
"eval_runtime": 2113.5402, |
|
"eval_samples_per_second": 1.35, |
|
"eval_steps_per_second": 0.169, |
|
"eval_wer": 42.6840255878908, |
|
"step": 3570 |
|
}, |
|
{ |
|
"epoch": 16.0, |
|
"grad_norm": 0.02384033426642418, |
|
"learning_rate": 3.6006051437216344e-06, |
|
"loss": 0.0002, |
|
"step": 3808 |
|
}, |
|
{ |
|
"epoch": 16.0, |
|
"eval_loss": 0.3772105872631073, |
|
"eval_runtime": 2110.6191, |
|
"eval_samples_per_second": 1.352, |
|
"eval_steps_per_second": 0.169, |
|
"eval_wer": 42.54887827732228, |
|
"step": 3808 |
|
}, |
|
{ |
|
"epoch": 17.0, |
|
"grad_norm": 0.05735332891345024, |
|
"learning_rate": 3.2005379055303416e-06, |
|
"loss": 0.0001, |
|
"step": 4046 |
|
}, |
|
{ |
|
"epoch": 17.0, |
|
"eval_loss": 0.3818971514701843, |
|
"eval_runtime": 2115.4194, |
|
"eval_samples_per_second": 1.349, |
|
"eval_steps_per_second": 0.169, |
|
"eval_wer": 42.64798630507253, |
|
"step": 4046 |
|
}, |
|
{ |
|
"epoch": 18.0, |
|
"grad_norm": 0.007810983806848526, |
|
"learning_rate": 2.8004706673390488e-06, |
|
"loss": 0.0001, |
|
"step": 4284 |
|
}, |
|
{ |
|
"epoch": 18.0, |
|
"eval_loss": 0.38705670833587646, |
|
"eval_runtime": 2106.3665, |
|
"eval_samples_per_second": 1.354, |
|
"eval_steps_per_second": 0.169, |
|
"eval_wer": 42.71555996035679, |
|
"step": 4284 |
|
}, |
|
{ |
|
"epoch": 19.0, |
|
"grad_norm": 0.017347875982522964, |
|
"learning_rate": 2.4004034291477564e-06, |
|
"loss": 0.0001, |
|
"step": 4522 |
|
}, |
|
{ |
|
"epoch": 19.0, |
|
"eval_loss": 0.3907012939453125, |
|
"eval_runtime": 2108.1964, |
|
"eval_samples_per_second": 1.353, |
|
"eval_steps_per_second": 0.169, |
|
"eval_wer": 42.67952067753852, |
|
"step": 4522 |
|
}, |
|
{ |
|
"epoch": 20.0, |
|
"grad_norm": 0.014235159382224083, |
|
"learning_rate": 2.000336190956463e-06, |
|
"loss": 0.0001, |
|
"step": 4760 |
|
}, |
|
{ |
|
"epoch": 20.0, |
|
"eval_loss": 0.3938260078430176, |
|
"eval_runtime": 2109.3918, |
|
"eval_samples_per_second": 1.353, |
|
"eval_steps_per_second": 0.169, |
|
"eval_wer": 42.6840255878908, |
|
"step": 4760 |
|
}, |
|
{ |
|
"epoch": 21.0, |
|
"grad_norm": 0.020930150523781776, |
|
"learning_rate": 1.6002689527651708e-06, |
|
"loss": 0.0001, |
|
"step": 4998 |
|
}, |
|
{ |
|
"epoch": 21.0, |
|
"eval_loss": 0.3964875340461731, |
|
"eval_runtime": 2107.6968, |
|
"eval_samples_per_second": 1.354, |
|
"eval_steps_per_second": 0.169, |
|
"eval_wer": 42.67501576718623, |
|
"step": 4998 |
|
}, |
|
{ |
|
"epoch": 22.0, |
|
"grad_norm": 0.01044421922415495, |
|
"learning_rate": 1.2002017145738782e-06, |
|
"loss": 0.0001, |
|
"step": 5236 |
|
}, |
|
{ |
|
"epoch": 22.0, |
|
"eval_loss": 0.398556649684906, |
|
"eval_runtime": 2118.6844, |
|
"eval_samples_per_second": 1.347, |
|
"eval_steps_per_second": 0.169, |
|
"eval_wer": 42.638976484367966, |
|
"step": 5236 |
|
}, |
|
{ |
|
"epoch": 23.0, |
|
"grad_norm": 0.01874064840376377, |
|
"learning_rate": 8.001344763825854e-07, |
|
"loss": 0.0, |
|
"step": 5474 |
|
}, |
|
{ |
|
"epoch": 23.0, |
|
"eval_loss": 0.40004006028175354, |
|
"eval_runtime": 2115.6844, |
|
"eval_samples_per_second": 1.348, |
|
"eval_steps_per_second": 0.169, |
|
"eval_wer": 42.638976484367966, |
|
"step": 5474 |
|
}, |
|
{ |
|
"epoch": 24.0, |
|
"grad_norm": 0.011713994666934013, |
|
"learning_rate": 4.000672381912927e-07, |
|
"loss": 0.0, |
|
"step": 5712 |
|
}, |
|
{ |
|
"epoch": 24.0, |
|
"eval_loss": 0.4009442925453186, |
|
"eval_runtime": 2114.295, |
|
"eval_samples_per_second": 1.349, |
|
"eval_steps_per_second": 0.169, |
|
"eval_wer": 42.620956842958826, |
|
"step": 5712 |
|
}, |
|
{ |
|
"epoch": 25.0, |
|
"grad_norm": 0.020036796107888222, |
|
"learning_rate": 0.0, |
|
"loss": 0.0, |
|
"step": 5950 |
|
}, |
|
{ |
|
"epoch": 25.0, |
|
"eval_loss": 0.40132761001586914, |
|
"eval_runtime": 2107.6565, |
|
"eval_samples_per_second": 1.354, |
|
"eval_steps_per_second": 0.169, |
|
"eval_wer": 42.607442111901975, |
|
"step": 5950 |
|
} |
|
], |
|
"logging_steps": 500, |
|
"max_steps": 5950, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 25, |
|
"save_steps": 500, |
|
"total_flos": 2.7372325220352e+19, |
|
"train_batch_size": 16, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|