search

adelacvg / ttts

Train the next generation of TTS systems.
Mozilla Public License 2.0
159 stars 17 forks source link

train hifigan failed #17

Closed wflrz123 closed 4 months ago

wflrz123 commented 6 months ago

I am currently retraining a 16k HiFi-GAN, but using latent as the input, the loss is difficult to converge. Although human voice can be heard, there is a presence of electronic sound. What could be the reason for this? Looking forward to your reply, thank you.

howitry commented 5 months ago

I am currently retraining a 16k HiFi-GAN, but using latent as the input, the loss is difficult to converge. Although human voice can be heard, there is a presence of electronic sound. What could be the reason for this? Looking forward to your reply, thank you.

Hello, I also encountered this problem when training 24k hifigan. Do you have a solution now?