[zero_to_fp32] fix padding removal #1380
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Outdated suggestions cannot be applied.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.
It appears that my initial implementation of zero_to_fp32 with zero3 was broken when padding was involved and it just happened to work since all models I worked with were w/o the padding (i.e. all params were even/divisible by n_gpus). This PR remedies that and adds a very specific test that verifies the various edges when params are not divisible by n_gpus.
@tjruwase