|
{ |
|
"best_metric": 0.2908598780632019, |
|
"best_model_checkpoint": "/models/hfhub/DewiBrynJones/wav2vec2-xlsr-53-ft-ccv-en-cy/checkpoint-9000", |
|
"epoch": 4.545454545454545, |
|
"eval_steps": 500, |
|
"global_step": 9000, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.2, |
|
"learning_rate": 0.00014925, |
|
"loss": 5.8377, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 0.25, |
|
"eval_loss": 1.2190359830856323, |
|
"eval_runtime": 1220.4523, |
|
"eval_samples_per_second": 21.545, |
|
"eval_steps_per_second": 0.674, |
|
"eval_wer": 0.8568814654720188, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 0.4, |
|
"learning_rate": 0.00029925, |
|
"loss": 0.9829, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 0.51, |
|
"eval_loss": 0.5584714412689209, |
|
"eval_runtime": 1188.1001, |
|
"eval_samples_per_second": 22.132, |
|
"eval_steps_per_second": 0.692, |
|
"eval_wer": 0.4700631596617496, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 0.61, |
|
"learning_rate": 0.00028543902439024386, |
|
"loss": 0.45, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 0.76, |
|
"eval_loss": 0.4735090434551239, |
|
"eval_runtime": 1202.8188, |
|
"eval_samples_per_second": 21.861, |
|
"eval_steps_per_second": 0.683, |
|
"eval_wer": 0.39008750026229094, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 0.81, |
|
"learning_rate": 0.00027080487804878047, |
|
"loss": 0.3636, |
|
"step": 1600 |
|
}, |
|
{ |
|
"epoch": 1.01, |
|
"learning_rate": 0.00025617073170731703, |
|
"loss": 0.3151, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 1.01, |
|
"eval_loss": 0.4124945402145386, |
|
"eval_runtime": 1160.0644, |
|
"eval_samples_per_second": 22.667, |
|
"eval_steps_per_second": 0.709, |
|
"eval_wer": 0.34183435801666073, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 1.21, |
|
"learning_rate": 0.00024153658536585365, |
|
"loss": 0.2524, |
|
"step": 2400 |
|
}, |
|
{ |
|
"epoch": 1.26, |
|
"eval_loss": 0.3830628991127014, |
|
"eval_runtime": 1409.4338, |
|
"eval_samples_per_second": 18.656, |
|
"eval_steps_per_second": 0.583, |
|
"eval_wer": 0.3117191598296158, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 1.41, |
|
"learning_rate": 0.00022690243902439024, |
|
"loss": 0.243, |
|
"step": 2800 |
|
}, |
|
{ |
|
"epoch": 1.52, |
|
"eval_loss": 0.3661448061466217, |
|
"eval_runtime": 1425.6726, |
|
"eval_samples_per_second": 18.444, |
|
"eval_steps_per_second": 0.577, |
|
"eval_wer": 0.30780368046666806, |
|
"step": 3000 |
|
}, |
|
{ |
|
"epoch": 1.62, |
|
"learning_rate": 0.0002122682926829268, |
|
"loss": 0.2341, |
|
"step": 3200 |
|
}, |
|
{ |
|
"epoch": 1.77, |
|
"eval_loss": 0.3488619327545166, |
|
"eval_runtime": 1416.3145, |
|
"eval_samples_per_second": 18.566, |
|
"eval_steps_per_second": 0.58, |
|
"eval_wer": 0.2883060201019787, |
|
"step": 3500 |
|
}, |
|
{ |
|
"epoch": 1.82, |
|
"learning_rate": 0.00019763414634146341, |
|
"loss": 0.2235, |
|
"step": 3600 |
|
}, |
|
{ |
|
"epoch": 2.02, |
|
"learning_rate": 0.00018299999999999998, |
|
"loss": 0.211, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 2.02, |
|
"eval_loss": 0.34998372197151184, |
|
"eval_runtime": 1154.299, |
|
"eval_samples_per_second": 22.78, |
|
"eval_steps_per_second": 0.712, |
|
"eval_wer": 0.27384434605619323, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 2.22, |
|
"learning_rate": 0.0001683658536585366, |
|
"loss": 0.1702, |
|
"step": 4400 |
|
}, |
|
{ |
|
"epoch": 2.27, |
|
"eval_loss": 0.34594303369522095, |
|
"eval_runtime": 1230.9263, |
|
"eval_samples_per_second": 21.362, |
|
"eval_steps_per_second": 0.668, |
|
"eval_wer": 0.27035272887508655, |
|
"step": 4500 |
|
}, |
|
{ |
|
"epoch": 2.42, |
|
"learning_rate": 0.00015373170731707315, |
|
"loss": 0.1634, |
|
"step": 4800 |
|
}, |
|
{ |
|
"epoch": 2.53, |
|
"eval_loss": 0.33047276735305786, |
|
"eval_runtime": 1138.5459, |
|
"eval_samples_per_second": 23.095, |
|
"eval_steps_per_second": 0.722, |
|
"eval_wer": 0.2583125249176406, |
|
"step": 5000 |
|
}, |
|
{ |
|
"epoch": 2.63, |
|
"learning_rate": 0.00013909756097560974, |
|
"loss": 0.1608, |
|
"step": 5200 |
|
}, |
|
{ |
|
"epoch": 2.78, |
|
"eval_loss": 0.3136747479438782, |
|
"eval_runtime": 1199.9023, |
|
"eval_samples_per_second": 21.914, |
|
"eval_steps_per_second": 0.685, |
|
"eval_wer": 0.24792580313490148, |
|
"step": 5500 |
|
}, |
|
{ |
|
"epoch": 2.83, |
|
"learning_rate": 0.00012446341463414633, |
|
"loss": 0.1542, |
|
"step": 5600 |
|
}, |
|
{ |
|
"epoch": 3.03, |
|
"learning_rate": 0.00010982926829268292, |
|
"loss": 0.1481, |
|
"step": 6000 |
|
}, |
|
{ |
|
"epoch": 3.03, |
|
"eval_loss": 0.32883504033088684, |
|
"eval_runtime": 1167.9676, |
|
"eval_samples_per_second": 22.513, |
|
"eval_steps_per_second": 0.704, |
|
"eval_wer": 0.2562183939400298, |
|
"step": 6000 |
|
}, |
|
{ |
|
"epoch": 3.23, |
|
"learning_rate": 9.519512195121951e-05, |
|
"loss": 0.1216, |
|
"step": 6400 |
|
}, |
|
{ |
|
"epoch": 3.28, |
|
"eval_loss": 0.3174259662628174, |
|
"eval_runtime": 1161.665, |
|
"eval_samples_per_second": 22.636, |
|
"eval_steps_per_second": 0.708, |
|
"eval_wer": 0.24461464213022222, |
|
"step": 6500 |
|
}, |
|
{ |
|
"epoch": 3.43, |
|
"learning_rate": 8.05609756097561e-05, |
|
"loss": 0.1181, |
|
"step": 6800 |
|
}, |
|
{ |
|
"epoch": 3.54, |
|
"eval_loss": 0.30002185702323914, |
|
"eval_runtime": 1207.6988, |
|
"eval_samples_per_second": 21.773, |
|
"eval_steps_per_second": 0.681, |
|
"eval_wer": 0.2324569318253352, |
|
"step": 7000 |
|
}, |
|
{ |
|
"epoch": 3.64, |
|
"learning_rate": 6.592682926829267e-05, |
|
"loss": 0.1143, |
|
"step": 7200 |
|
}, |
|
{ |
|
"epoch": 3.79, |
|
"eval_loss": 0.29287537932395935, |
|
"eval_runtime": 1095.1242, |
|
"eval_samples_per_second": 24.011, |
|
"eval_steps_per_second": 0.751, |
|
"eval_wer": 0.23258702813857357, |
|
"step": 7500 |
|
}, |
|
{ |
|
"epoch": 3.84, |
|
"learning_rate": 5.129268292682927e-05, |
|
"loss": 0.1139, |
|
"step": 7600 |
|
}, |
|
{ |
|
"epoch": 4.04, |
|
"learning_rate": 3.665853658536585e-05, |
|
"loss": 0.1049, |
|
"step": 8000 |
|
}, |
|
{ |
|
"epoch": 4.04, |
|
"eval_loss": 0.2921387255191803, |
|
"eval_runtime": 1091.9168, |
|
"eval_samples_per_second": 24.082, |
|
"eval_steps_per_second": 0.753, |
|
"eval_wer": 0.22178483748452482, |
|
"step": 8000 |
|
}, |
|
{ |
|
"epoch": 4.24, |
|
"learning_rate": 2.2024390243902437e-05, |
|
"loss": 0.0913, |
|
"step": 8400 |
|
}, |
|
{ |
|
"epoch": 4.29, |
|
"eval_loss": 0.2968423664569855, |
|
"eval_runtime": 1104.263, |
|
"eval_samples_per_second": 23.812, |
|
"eval_steps_per_second": 0.744, |
|
"eval_wer": 0.2208196067734016, |
|
"step": 8500 |
|
}, |
|
{ |
|
"epoch": 4.44, |
|
"learning_rate": 7.4268292682926826e-06, |
|
"loss": 0.0883, |
|
"step": 8800 |
|
}, |
|
{ |
|
"epoch": 4.55, |
|
"eval_loss": 0.2908598780632019, |
|
"eval_runtime": 1095.7607, |
|
"eval_samples_per_second": 23.997, |
|
"eval_steps_per_second": 0.75, |
|
"eval_wer": 0.21777283505046477, |
|
"step": 9000 |
|
}, |
|
{ |
|
"epoch": 4.55, |
|
"step": 9000, |
|
"total_flos": 8.920242900072627e+19, |
|
"train_loss": 0.47586327913072374, |
|
"train_runtime": 55725.2112, |
|
"train_samples_per_second": 10.336, |
|
"train_steps_per_second": 0.162 |
|
} |
|
], |
|
"logging_steps": 400, |
|
"max_steps": 9000, |
|
"num_train_epochs": 5, |
|
"save_steps": 500, |
|
"total_flos": 8.920242900072627e+19, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|