scikit-learn / scikit-learn/scikit-learn
Threads awaiting for GIL in Forest estimators
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 67.3k
- Forks
- 27.4k
- Avg merge
- 1d 15h
- Merged PRs (30d)
- 58
Description
Discussed in https://github.com/scikit-learn/scikit-learn/discussions/20651
In forest algorithms, the preferred parallelization backend is threading. However, it looks that it is not anymore the most appropriate backend. As discussed here, it might be that the GIL is not explicitly released in some part of the code locking the execution of the thread.
We need to investigate more to solve this issue.
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 reading discussion #20651, especially the linked comment, then reproduce the forest-estimator behavior with the threading backend and investigate where threads wait for the GIL. The issue names no files or tests, so use the estimator parallelization paths as the entry point. Done means identifying the blocking code and resolving the backend or GIL-release problem.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- machine-learning, performance
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100