Refine ca #1263
Merged
Refine ca #1263
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Contributor
There was a problem hiding this comment.
Pull Request Overview
This PR refactors the custom all-reduce API by consolidating separate "registered" and "unregistered" variants into a unified interface. The main changes include renaming full_nvlink to fully_connected for better clarity and merging all_reduce_reg and all_reduce_unreg into a single all_reduce function that accepts an optional registration buffer parameter.
Key changes:
- Unified all-reduce API with optional registration buffer parameter
- Renamed parameter from
full_nvlinktofully_connectedfor improved clarity - Import statement reorganization for consistent code style
Reviewed Changes
Copilot reviewed 9 out of 15 changed files in this pull request and generated 1 comment.
Show a summary per file
| File | Description |
|---|---|
| op_tests/multigpu_tests/test_custom_allreduce.py | Reorganized imports to follow standard conventions (standard library, third-party, local imports) |
| csrc/kernels/custom_all_reduce.cu | Merged all_reduce_reg and all_reduce_unreg into single all_reduce function with optional reg_buffer parameter; renamed full_nvlink to fully_connected |
| csrc/include/rocm_ops.hpp | Updated Python bindings to reflect unified all_reduce API and renamed parameter |
| csrc/include/custom_all_reduce.h | Updated function signatures for unified API and improved formatting |
| csrc/include/custom_all_reduce.cuh | Renamed constructor parameter from full_nvlink to fully_connected |
| aiter/ops/sample.py | Reorganized imports and renamed temperatures parameter to temperature (singular) |
| aiter/ops/custom_all_reduce.py | Updated Python interface to match unified C++ API |
| aiter/ops/communication.py | Updated references to access custom all-reduce communicator through device_communicator |
| aiter/jit/core.py | Removed unused operation from export list |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
junhaha666
approved these changes
Oct 27, 2025
ganyi1996ppo
pushed a commit
that referenced
this pull request
Nov 19, 2025
* refine CA * fix * fix type hint
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
Motivation
Technical Details
Test Plan
Test Result
Submission Checklist