search

yl4579 / StyleTTS2

StyleTTS 2: Towards Human-Level Text-to-Speech through Style Diffusion and Adversarial Training with Large Speech Language Models
MIT License
5k stars 423 forks source link

Update training scripts to use generic device #229

Open ffd000 opened 7 months ago

ffd000 commented 7 months ago

Updated the training scripts to use specified device instead of moving stuff to cuda exclusively.

Faolain commented 2 months ago

@yl4579 is this repository still maintained?

78Alpha commented 2 months ago

@yl4579 is this repository still maintained?

Most advancement is on the dagshub repo. This appears to be a research only repo.