-
Notifications
You must be signed in to change notification settings - Fork 3
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
ERROR:Failed to load the extension "edge_tts". [...] ValueError: mutable default <class 'fairseq.dataclass.configs.CommonConfig'> for field common is not allowed: use default_factory #4
Comments
Never seen this issue before. Ensure you install the requirements before running text-generation-webui.
|
I am having the same issue, fresh install of text generation webui and also installed requirements
|
same issue last line in intslal requzrement.txt if i start web-ui |
For now delete the then edit the double click the the issue is python 3.11. I can confirm this will fix it. |
sry same issue: (e:\text-generation-webui\installer_files\env) e:\text-generation-webui>python 2023-11-14 16:39:38 ERROR:Failed to load the extension "edge_tts". |
Run it should work now, the package was not installed correctly because of the incompatibility of the modules and 3.11. Let me know if it works |
THX !!! do you know a software to generate your own rvc ? and may if i can reduce the speed of the voice that would me great |
Actually, I have this one that is from FNAF phone guy, it sounds pretty cool to me. Has this fallout 4 vibe to it: I use this github project for generating RVC and Bark voice clones, its like text generation webui. It's seriously underrated: For RVC models I typically get them from a discord channel here is the invite link: They have a spreadsheet with a bunch of voices : https://docs.google.com/spreadsheets/d/1leF7_c2Qf5iQRVkmOF51ZSynOvEjz8fHqwriX1wUMPw/edit#gid=0 Have fun! |
omg 24h service ^^ thx , so no noises, like breath or snore or laughing ? |
No problem ! You need bark ai for that, it sounds more natural. I think there's a bark extension for text generation webui but I am not sure if it has rvc included. You might have to create something custom. I suppose you can use some of the code from here however the only downside is that bark takes longer to generate audio since the response has to be sliced in pieces, processed and then stitched back up in audio before you can hear the whole script and not to mention its not consistent at times. if i am not mistaken the index file is part of rvc 2 that goes with the pt files. I am new to voice cloning too and still learning |
i see, i take look ... maybe one question for the moment... |
There's a settings-template you can use in the same directory of start_windows. Remove the template part and it will load automatically. You can also load models automatically using the CMD txt file. I would check oobabooga repo for more arguments, they support a bunch. |
i will take a look omg, I have noticed that edge_tts must work online ... why ? |
Seems the tts is a Microsoft online service, not sure if we can bypass it |
it downlaod at least 1mb every time it speaks, and I don't want to know if anything will be uploaded ^^ can you ask if the programmers can proofe it ? |
@BuffMcBigHuge I am not sure, maybe BuffMcBigHuge knows more. You can try this bark extension https://github.com/wsippel/bark_tts |
Sadly, getting the same problems. Running in Py3.10/Admin priviligies didnt helped. Also tried different versions of libraries, also didnt helped. Running newest version of Ooba. Any ideas? |
I may have fixed the issue. It's related to fairseq not being supported with Python 3.11, which is now default in the conda environment that text-generation-webui creates with the start bat/sh files. I modified the requirements to point to a repo fork of fairseq that supports 3.11. You may thank Facebook for this. 🤭 |
The text was updated successfully, but these errors were encountered: