Skip to content
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

Debug ST commit (this time it works) #109

Merged
merged 11 commits into from
Jan 15, 2025
Merged

Conversation

Sylphar
Copy link
Contributor

@Sylphar Sylphar commented Jan 15, 2025

Main goal was to stop that annoying problem when the user gets a multi-paragraph answer. Before, the issue with this was that SillyTavern breaks up its paragraphs internally with

. This made the script crash instantly sinceit couldn't distinguish what the body of the message was, I think ? So, I tried to make it simply intake all the

in the mes_last category.
This highlighted another issue after a good amount of debugging (with a new, better testAdvST.py) : after a few messages, ST gets a bit slow, and it was too easy for the script to see the generate button before it was fully pressed/changed to Stop, causing the user to get back their own input as output. This is fixed with a small delay and a wait_for_selector of the stop button, like in Text-gen.

Never encountered those with my test bot Testy... they cause strict mode violation: locator(".mes.last_mes .mes_text p") resolving to multiple elements, crashing MonikAI. This should fix it ?
Just allows the script to act as if it were still in the main directory, just an operation to bring the working directory one subfolder up.
+ same logic change as main.py
See TestAdvST, also just remove all common warnings completely they're annoying and not very useful
See TestAdvST
They're just annoying and useless for most users. So, I removed UserWarnings and FutureWarnings. The people who'd be interested in these info probably know how to open the file and see for themselves that they have been suppressed.
Small, easy to distinguish voice file that works with the test scripts. That is all.
Avoids the script looking at the generate button while it's in the process of turning into the stop button.
See main.py. Also, added more debug states to be detected, and an option to just type in whatever you want.
See main.py
Since it's a common complaint I sometimes hear that you can't get long messages before the buffer gets overwhelmed. Bumped it to 2048
Copy link
Owner

@Rubiksman78 Rubiksman78 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Looks good to me !

@Rubiksman78 Rubiksman78 merged commit a6b4abc into Rubiksman78:main Jan 15, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants