Contribution: "Daniela" Voice Trained with Piper on OpenSLR Dataset
#35
by
larcanio
- opened
Hi,
I've trained a new Argentinian Spanish voice model for Piper named "Daniela", using the OpenSLR 61 dataset.
The model is available here: https://huggingface.co/larcanio/piper-voices
Please let me know if you'd be interested in including it in the official voice collection.
Thanks again for all the great work on the project!
Best regards,
Laureano
Absolutely, thank you for the contribution! I've added it into the official voice collection :)
Hi there, please let me know where would be a good place to put a script that converts an slr zip file to one that can be easily used to train voices for piper.
- goes from the tsv format to the _transcriptions.txt file and
- downsample files to 22.050Hz.
It can be usable with the notebooks.
BTW, Daniela sounds great, latam localizing is really interesting. Thanks for all the work