meta-pytorch / meta-pytorch/data
[Stateful DataLoader] add utilities to inspect/modify dataloader state
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 1.3k
- Forks
- 179
- Avg merge
- 6d 1h
- Merged PRs (30d)
- 2
Description
🚀 The feature
State Dict utilities to inspect state
Motivation, pitch
We should provide some utilites for users to inspect and potentially modify the dataloader state. Currently the format of the dictionary is opaque, and we may want to make changes in the future. Instead of users inspecting and hard-coding paths, we should instead provide utilities to make it easy, and future-proof.
Current list of desired utilities:
- Retrieve state for worker 'i'
- Retrieve and update the number of steps to skip upon resume
Alternatives
No response
Additional context
No response
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
No files, tests, or entry points are named. Start by locating the Stateful DataLoader state-dict implementation and its existing serialization tests, then determine the public API needed to retrieve a worker's state and get or update resume-step counts. Done means both utilities are defined with behavior and compatibility expectations that are clear from tests or documentation.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- data
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 30/100