deepspeedai / deepspeedai/DeepSpeed
Exception with optimization_stage_3, now in backward
Open
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 43.1k
- Forks
- 5k
- Avg merge
- 4d 15h
- Merged PRs (30d)
- 112
Description
Following up on issue #944 for which I have submitted code to reproduce here, the repro code goes through forward, but now fails on backward with a different exception:
It would be great if the repro code could be tried with potential fixes so that this issue can be fully addressed this time.
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 with the reproduction notebook linked in the issue and compare its forward and backward paths under optimization_stage_3. Review issue #944 for prior context, then run the repro to capture the backward exception and test potential fixes. Done means the repro completes backward without the reported exception.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python, pytorch
- Domain
- distributed-systems, machine-learning
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100