combine dp and stage2 hybrid parallel. #46795
Merged
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
PR types
New features
PR changes
Others
Describe
combine dp and stage2 hybrid parallel
in the past, we can not use sharding stage2 and data parallel togethor,this PR combines dp and stage2 hybrid communication. users just only pass a dp_group param into group_sharded_parallel if they want to hybrid dp and sharding stage2.
we compare the loss diff on 1.3B GPT model as shown below. [dp8] VS [dp2 + sharding4]