-
Notifications
You must be signed in to change notification settings - Fork 58
optimize _update_per_bucket_p2p logic #28
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
base: main
Are you sure you want to change the base?
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Pull Request Overview
This PR optimizes the P2P (peer-to-peer) update logic by merging the _update_per_bucket_p2p function with the existing _update_per_bucket function and implementing a topology-aware receiver assignment strategy to maximize bandwidth utilization.
- Consolidated P2P and broadcast update logic into a single unified function
- Added RDMA device topology awareness for optimal receiver assignment
- Enhanced metadata structures to include RDMA device information
Comments suppressed due to low confidence (1)
checkpoint_engine/ps.py:1
- This line was removed but
self._current_global_parameter_metasis still being used later in the function without proper initialization. This could cause issues if the dictionary is not properly initialized elsewhere.
from __future__ import annotations
Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.
Signed-off-by: specture724 <149605198+specture724@users.noreply.github.com>
Signed-off-by: specture724 <149605198+specture724@users.noreply.github.com>
…-engine into feat/optimize_p2p
Resolve #25
Merge _update_per_bucket_p2p with _update_per_bucket. Topology-awared receiver assignment strategy implemented to make full use of bandwidth