Replies: 1 comment 2 replies
-
|
The speakers of the VCTK Vits model are randomly shuffled, see #2258, so can't be compared with any other VCTK models. |
Beta Was this translation helpful? Give feedback.
2 replies
-
|
That's why I made the repo comparing the available speakers, due to them all being shuffled. Are you saying they all get shuffled every time the model is loaded? |
Beta Was this translation helpful? Give feedback.
-
|
Ok, it wasn't clear to me from your post or that repo and because you wrote
|
Beta Was this translation helpful? Give feedback.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Inspired by: #1891
https://github.com/T145/vctk-vits-speaker-comparison
I tried to use
tts_models/en/vctk/fast_pitch, but the speaker IDs didn't work.Has anyone had similar results and preferences? Or any that are completely different?
Beta Was this translation helpful? Give feedback.
All reactions