[misc]: generate_and_rm_group is actually executed serially, without realizing true concurrency.
Open
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 8.5k
- Forks
- 1.3k
- Avg merge
- 5h 36m
- Merged PRs (30d)
- 22
Description
In the file sglang_rollout.py, the calculation of group rewards is actually executed serially, without realizing true concurrency.
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 in sglang_rollout.py and trace generate_and_rm_group to see how group rewards are currently calculated and where execution becomes serial. Confirm the intended concurrency behavior from the surrounding rollout flow; the issue is done when group reward calculations execute concurrently rather than serially.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- machine-learning
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 38/100