jaywalnut310 / vits

VITS: Conditional Variational Autoencoder with Adversarial Learning for End-to-End Text-to-Speech
https://jaywalnut310.github.io/vits-demo/index.html
MIT License
6.53k stars 1.22k forks source link

Question about strange voice #91

Open panxin801 opened 1 year ago

panxin801 commented 1 year ago

Hello, sir. Thank you for you sharing your great works. And after training my TTS model, I find a weird voice during inference .

I use chinese dataset for training, during interface there's a little tick voice in the middle of my sample.

企业微信20221020-193656@2x

As you can see I point it with white rectangle. Do you have a idea about how to fix it or any parts of the model may have problem? I'm a rookie in TTS. Thanks you for your time and advice.