-
Notifications
You must be signed in to change notification settings - Fork 7.4k
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
Gets stuck in infinte loops when requested to generate code #58
Comments
Also happens sometimes in Linux, like a loop of verbiage |
It is Said in the Training Log About This... |
There it produced valid code. Here there’s no valid code. No idea if the underlying issue is the same. |
There should be options |
It won't correctly work, at least in my example. It answers a different question.
Well, fair enough. Sounds like useful code generation is a bit too hard for it atm :-) |
It even tries to circumvent repeat penalty!
|
@AndriyMulyar Why is it closed without any notice? This bug still exists. |
Closed as this is not an issue that can be solved with a pull request. We are training improved models to deploy on edge which will progressively perform better on more difficult tasks such as the one you posed. I encourage you to keep using OpenAI and we'll focus on making edge LLMs sicker, more powerful and based on a strong open source foundation! |
I used ./gpt4all-lora-quantized-OSX-m1
Another session:
The text was updated successfully, but these errors were encountered: