mirror of
https://github.com/SillyTavern/SillyTavern-Extras.git
synced 2026-10-15 18:11:52 +00:00
[GH-ISSUE #217] Whisper Troubleshoot Request #133
Labels
No labels
bug
duplicate
planned
pull-request
under consideration
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference
SillyTavern/SillyTavern-Extras#133
Loading…
Add table
Add a link
Reference in a new issue
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?
Originally created by @Statford on GitHub (Jan 15, 2024).
Original GitHub issue: https://github.com/SillyTavern/SillyTavern-Extras/issues/217
Hello,
First off I want to apologize if this is a commonly asked question and if I have broken any of the rules in regards to looking first? I had a poke around to find my problem and this is my first time doing something like this.
I downloaded Whisper-stt via SIllyTavern Extras and I can't seem to get it to work. After a week and a half of reading documentation, looking up youtube videos, and googling for answers to my problem; I'm at a loss to how to fix it.
I had some people help me troubleshoot who had working copies and I've noticed that when I do boot up it never says "Initializing Whisper speech-recognition (from ST request file)" when I do boot up. Everything else in STExtras works fine, SillyTavern is up-to-date, and the Speech Recognition add-on is working as I have used stt via the browser to make sure it wasn't my microphone.
I'm fairly positive it has something to do with a borked Whisper install via SillyTavern Extras but I was hoping to get some more light shed on what happened so I can at least go to Whisper (If I have too) with more information.
Thanks!
@Cohee1207 commented on GitHub (Jan 16, 2024):
Have you read the docs? https://docs.sillytavern.app/extras/extensions/speech-recognition/
By screenshot it looks like you're missing it from the enable-modules string.
@Statford commented on GitHub (Jan 16, 2024):
I use a .bat to boot and this is what my .bat looks like:
cd C:\Users\jocat\SillyTavern-extras
call conda activate extras
python server.py --enable-modules=whisper-stt--cpu
call conda deactivate
pause
@Cohee1207 commented on GitHub (Jan 16, 2024):
You're missing a space between stt and --cpu
@Statford commented on GitHub (Jan 16, 2024):
Whoops, fixed that and I got this :/.
@Statford commented on GitHub (Jan 17, 2024):
I'm still troubleshooting this problem, I did some research and they said it might be a pathing problem to python? I can "where python" and find it but I'm kind of unsure how to fix it other than going line by line in every .py to see where the problem is.
@Statford commented on GitHub (Jan 18, 2024):
After testing some solutions out yesterday, I think the problem was ultimately user error :p in me poking around in the conda modules for Whisper or in ST-Extra's files. I've uninstalled and reninstalled ST-Extras but I still have the same error, does ST keep previously downloaded (and edited) documents when you uninstall? I have no idea how to uninstall and reinstall the conda end of Whisper.
@Cohee1207 commented on GitHub (Jan 18, 2024):
Delete the conda env and start anew.
https://stackoverflow.com/a/49128056
@Statford commented on GitHub (Jan 18, 2024):
Okay, I'll also reinstall ST-Extras on top of that. But I have a very embarrasing question, ST has to be in staging right?
@Statford commented on GitHub (Jan 21, 2024):
Okay, so I reinstalled everything and now Whisper is actually working on sillytavern but no matter what I say the output is always just "you".
@Statford commented on GitHub (Jan 27, 2024):
Okay, so I let the problem sit for a week before going back to it and now all of a sudden it works after reboot. I'm running it on release too, which surprized me because a lot of people were telling me that I should be in staging, but I also heard somewhere that you no longer need to use staging as it could be done on release as well, so maybe the doumentation for Whisper STT should be updated to reflect that?
Either way, I'm estatic that it now works but I do have 1 final question, if you would indulge me before I close out the thread.
What would you recommend I read (outside of the documentation that comes with the programs) or watch to help me get better at troubleshooting stuff like this in the future? Thanks!
@Statford commented on GitHub (Feb 9, 2024):
I assume you've been busy, thanks for the help! I appreciate it!