Trying the deepspeech voice-corpus-tools, I decided to augment my TRAIN dir, with tempo and speed (±5%).
How could I tell to deepspeech to use more than one csv file ?
I have 3 CSV files, with same wav names !!!
Thanks.
Trying the deepspeech voice-corpus-tools, I decided to augment my TRAIN dir, with tempo and speed (±5%).
How could I tell to deepspeech to use more than one csv file ?
I have 3 CSV files, with same wav names !!!
Thanks.
Thanks Reuben.
--train_files /first.csv, /next.csv \
just use some csv files, separed by a comma !
note : here, both csv files have same wav name (objective : duplicate my train voices with voice-corpus-tools, and improve accuracy with some minor changes in sounds)
works perfectly.