instruction_following environment quality verification
Open
Nobody has claimed this yet.
close-confirm-done
env-infra
- Dominant language
- Python
- Stars
- 1.2k
- Forks
- 349
- Avg merge
- 1d 21h
- Merged PRs (30d)
- 318
Description
Use cases, pain points, and background
instruction_following resources server is currently unverified
Description:
Design:
May need to implement IFEval and IFBench inside Gym
Out of scope:
Acceptance Criteria:
- A full W&B run where we see the IFEval and IFBench validation accuracy improve meaningfully
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by locating the instruction_following resources server and reviewing how environments and validation metrics are implemented in Gym. Then investigate the intended IFEval and IFBench integration points and the W&B run configuration. Done means a full W&B run shows meaningful improvement in both IFEval and IFBench validation accuracy.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- machine-learning, testing-qa
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100