deepspeedai / deepspeedai/DeepSpeedExamples
Megatron-LM-v1.1.5-ZeRO3: evaluation during training disabled
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 6.8k
- Forks
- 1.1k
- Avg merge
- 2d 16h
- Merged PRs (30d)
- 1
Description
In the Megatron-LM-v1.1.5-ZeRO3 Megatron implementation, periodic evaluations do not happen during training. This is essential functionality for knowing that training is proceeding correctly. Evaluation can be reenabled by removing the quotation marks around lines 484-491 of megatron/training.py. I have tested this with several DeepSpeed settings, and the evaluation seems to work properly. But those lines were clearly quoted out intentionally, since the comment above says # XXX temporarily disabled for ZeRO-3. It would be helpful to know the reason for this at least. Does evaluation work incorrectly only if the ZeRO Optimizer is set to stage 3? What would be needed to get evaluation working for the cases where it doesn't work?
Contributor guide
No contributing guide indexed for this repository
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 megatron/training.py around lines 484-491 and read the surrounding # XXX temporarily disabled for ZeRO-3 comment. Reproduce evaluation with the reported DeepSpeed configurations, then determine whether it is safe for ZeRO-3 and what change or explanation is needed for cases where it is not.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- distributed-systems, machine-learning
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 45/100