0.9.0 Download

Update content

  1. The newly trained model has higher quality and better emotional tone
  2. Automatic cleaning noise optimization, cleaning effect is better
  3. Train faster
  4. Increase speech speed adjustment during generation
  5. When downloading, the last selected directory is automatically selected. You do not need to select a location each time
  6. Solve the training failure caused by the training audio speed too fast
  7. Solve the problem that part of the audio is discarded as invalid audio during training