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.

Sign up for GitHub

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

Jump to bottom

OpenAI noted that the logprobs of their instruction tuned model was out of whack, maybe use (var)entropy from a base model? #61

Open

HenkPoley opened this issue Oct 10, 2024 · 0 comments

HenkPoley commented Oct 10, 2024 •

edited

Loading

Something that popped in my head from the GPT-4 technical(?) report.

Yes, I mean loading two models. Don't ask the base model about its opinions of chat housekeeping tokens.

HenkPoley mentioned this issue

Open

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment