search

keonlee9420 / PortaSpeech

PyTorch Implementation of PortaSpeech: Portable and High-Quality Generative Text-to-Speech
MIT License
332 stars 36 forks source link

Missing pretrained models #4

Closed sygi closed 3 years ago

sygi commented 3 years ago

Hi, the link to the pretrained models seems to lead to an empty directory. Could you fix?

khromandr commented 3 years ago

Yeah, could you please upload them!

keonlee9420 commented 3 years ago

Hi @sygi and @khromandr, the model is successfully trained and converged on LJSpeech. Due to additional optimization experiments not being introduced in the paper, it takes some time to compare them. I will update the repo with the pre-trained models and demo samples as soon as the best model is out, so please stay tuned!

keonlee9420 commented 3 years ago

All pre-trained models with audio samples for both normal and small models are updated. Please check them out :)