mrq/ai-voice-cloning
31
74
Fork 43
Code Issues 222 Pull Requests 5 Packages Projects Releases Wiki Activity

Random voice not working with CVVP? #154

New Issue
Closed
opened 2023-03-18 20:09:51 +00:00 by psammites · 2 comments
psammites commented 2023-03-18 20:09:51 +00:00
Copy Link

With CVVP Weight above zero and the 'random' voice selected:

[1/1] Generating line: enaindvajset
Loading voice: random with model e9bde4de
Requesting weighing against CVVP weight, but voice latents are missing some extra data. Please regenerate your voice latents with 'Slimmer voice latents' unchecked.
Generating autoregressive samples
Computing best candidates using CLVP
Transforming autoregressive outputs into audio..
Generating line took 8.718473434448242 seconds
Loading Voicefixer
Loaded Voicefixer

'Slimmer voice latents' is off, but obviously regenerating latents doesn't do anything when voice is set to 'random'.

With CVVP Weight above zero and the 'random' voice selected: ```` [1/1] Generating line: enaindvajset Loading voice: random with model e9bde4de Requesting weighing against CVVP weight, but voice latents are missing some extra data. Please regenerate your voice latents with 'Slimmer voice latents' unchecked. Generating autoregressive samples Computing best candidates using CLVP Transforming autoregressive outputs into audio.. Generating line took 8.718473434448242 seconds Loading Voicefixer Loaded Voicefixer ```` 'Slimmer voice latents' is off, but obviously regenerating latents doesn't do anything when voice is set to 'random'.
mrq commented 2023-03-18 22:10:41 +00:00
Owner
Copy Link

Not possible, as the CVVP requires auto_conds from processing the original audio. Random voices never possessed actual speech to compare the source against the CVVP to rank AR samples against.

Random latents are merely just generating random values and passing them through the AR/diffusion conditioners.

Not possible, as the CVVP requires [auto_cond](https://git.ecker.tech/mrq/tortoise-tts/src/branch/main/tortoise/api.py#L505)s from processing the original audio. Random voices never possessed actual speech to compare the source against the CVVP to rank AR samples against. Random latents are merely just [generating](https://git.ecker.tech/mrq/tortoise-tts/src/branch/main/tortoise/api.py#L529) random values and passing them through the AR/diffusion conditioners.
mrq closed this issue 2023-03-18 22:10:41 +00:00
psammites commented 2023-03-19 01:48:16 +00:00
Author
Copy Link

Yes, what I meant was the error message is incorrect.

Yes, what I meant was the error message is incorrect.
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels   
bug

Something is not working

  
duplicate

This issue or pull request already exists

  
enhancement

New feature

  
help wanted

Need some help

  
insufficient info

Need more information

  
invalid

Something is wrong

  
news

To make up for a lack of a way to provide any news-like

  
not a bug

For non-bug things

  
question

More information is needed

  
wontfix

This won't be fixed

No Label
bug
duplicate
enhancement
help wanted
insufficient info
invalid
news
not a bug
question
wontfix
Milestone
Clear milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
No Assignees
2 Participants
Notifications
Due Date
The due date is invalid or out of range. Please use the format 'yyyy-mm-dd'.

No due date set.

Dependencies

No dependencies set.

Reference: mrq/ai-voice-cloning#154
No description provided.
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?