search

alphacep / vosk-tts

Text To Speech Synthesis with Vosk
Apache License 2.0
105 stars 15 forks source link

Audio hours in the dataset #24

Open Bubarinokk opened 2 months ago

Bubarinokk commented 2 months ago

How many hours of audio were used to make the Russian language model in GPT-SoVITS?

nshmyrev commented 2 months ago

about 1000

Bubarinokk commented 2 months ago

Do you have a method for finding several hours of audio in a language?

nshmyrev commented 2 months ago

Yes, sure, we can build any language you need.

Bubarinokk commented 2 weeks ago

Yes, sure, we can build any language you need.

How?

nshmyrev commented 2 weeks ago

You need to tell the language name

Bubarinokk commented 2 weeks ago

You need to tell the language name

Brazilian Portuguese

nshmyrev commented 1 week ago

@Bubarinokk like this https://github.com/Edresson/TTS-Portuguese-Corpus?

Bubarinokk commented 1 week ago

@Bubarinokk like this https://github.com/Edresson/TTS-Portuguese-Corpus?

Yes