Skip to content
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

Exclude non-training clients from cse #2400

Merged
merged 1 commit into from
Mar 15, 2024
Merged

Conversation

yanchengnv
Copy link
Collaborator

Fixes #4562056

Description

The CCWF's cross-site-eval workflow (CSE) assumes all evaluatees have local models. However, depending on the config of the Swarm Learning workflow (usually before CSE) of the job, some clients may only be aggregators and don't train. In this case, the client that is not a trainer won't be able to submit local model, which causes the CSE to fail.

This PR fixes this issue by checking the client against the training clients list of Swarm Learning workflow. Only training clients will be treated as evaluatees.

Types of changes

  • Non-breaking change (fix or new feature that would not break existing functionality).
  • Breaking change (fix or new feature that would cause existing functionality to change).
  • New tests added to cover the changes.
  • Quick tests passed locally by running ./runtest.sh.
  • In-line docstrings updated.
  • Documentation updated.

Copy link
Collaborator

@SYangster SYangster left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Nice fix, LGTM

@yanchengnv yanchengnv merged commit 08c202f into NVIDIA:2.4 Mar 15, 2024
15 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

2 participants