[PyTorch] Userbuffers support in operation-based API #1142
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Outdated suggestions cannot be applied.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.
Description
This PR adds basic support in the linear operation for using Userbuffers to overlap tensor-parallel communication with GEMMs. This is implemented as fused operations:
I've tried to avoid touching the core UB infrastructure in
transformer_engine/pytorch/module/base.py
, so I've kept the messy API and hackily worked around some bugs. This feature should be considered experimental and unstable.Type of change
Changes
Checklist: