Skip to content

Feat: more robust naive agent #13

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
Merged
merged 2 commits into from
Dec 17, 2024
Merged

Feat: more robust naive agent #13

merged 2 commits into from
Dec 17, 2024

Conversation

@DavidePaglieri
Copy link
Contributor

@DavidePaglieri DavidePaglieri commented Dec 17, 2024

A common problem with BALROG is that models that are weaker at instruction following will struggle to output the action as required in the naive.py agent.

We create a new more robust naive zero-shot agent, that is prompted with the following:

You must choose exactly one of the listed actions and output it strictly in the following format:
<|ACTION|>YOUR_CHOSEN_ACTION</|ACTION|>
You must not output any other text before or after these tags. No explanation, no reasoning, just the action within these tags.

@DavidePaglieri DavidePaglieri merged commit 9c65dad into main Dec 17, 2024
4 checks passed
@DavidePaglieri DavidePaglieri deleted the feat/robust_naive branch December 18, 2024 15:44
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@BartekCupial BartekCupial Awaiting requested review from BartekCupial

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

2 participants

@DavidePaglieri