Closed Colinsnow1 closed 6 months ago
Hello liuyangwen.
I'm currently pausing work on the Bert-VITS2 project. Based on my assessment of the fishaudio repository, it seemed challenging to train a Korean model there. Therefore, I am exploring methods with other VITS-based models using Korean data.
At the moment, I'm researching a VITS-based zero-shot multi-speaker Korean TTS model. If this model is successfully trained, I plan to incorporate BERT into it to ultimately create a BERT-VITS2 Korean model.
Thank you.
Dear DaegyeomKim,
I hope you are having a wonderful day. Thank you for your reply and wish you success in training!
Best regards.
Hello DaegyeomKim,
I am also trying to training a Korean version of Bert-VITS2. Have you training the Korean models successfully?
Thank you for sharing your excellent code. Have a great day.
Best regards.