Skip to content

Support allreduce fusion with cp#21249

Merged
Fridge003 merged 3 commits intosgl-project:mainfrom
Shunkangz:fix_allreduce_fusion_cp
Apr 20, 2026
Merged

Support allreduce fusion with cp#21249
Fridge003 merged 3 commits intosgl-project:mainfrom
Shunkangz:fix_allreduce_fusion_cp

Conversation

@Shunkangz
Copy link
Copy Markdown
Contributor

@Shunkangz Shunkangz commented Mar 24, 2026

Motivation

Enable the allreduce_fusion with context parallel. This requires the allreduce_fusion can work with any sub comm group.

Modifications

This PR can only work after flashInfer PR merged.

Accuracy Tests

Benchmarking and Profiling

Checklist

Review Process

  1. Ping Merge Oncalls to start the PR flow. See the PR Merge Process.
  2. Get approvals from CODEOWNERS and other reviewers.
  3. Trigger CI tests with comments or contact authorized users to do so.
    • /tag-run-ci-label, /rerun-failed-ci, /tag-and-rerun-ci
  4. After green CI and required approvals, ask Merge Oncalls to merge.

@gemini-code-assist
Copy link
Copy Markdown
Contributor

Warning

You have reached your daily quota limit. Please wait up to 24 hours and I will start processing your requests again!

@Fridge003
Copy link
Copy Markdown
Collaborator

/tag-and-rerun-ci

@Shunkangz
Copy link
Copy Markdown
Contributor Author

/tag-and-rerun-ci

Comment thread python/sglang/srt/layers/communicator.py
Comment thread python/sglang/srt/server_args.py
Comment thread python/sglang/srt/model_executor/model_runner.py
Comment thread python/sglang/srt/layers/flashinfer_comm_fusion.py
Comment thread python/sglang/srt/layers/flashinfer_comm_fusion.py
Comment thread python/sglang/srt/layers/flashinfer_comm_fusion.py
Comment thread python/sglang/srt/layers/flashinfer_comm_fusion.py
@Shunkangz Shunkangz force-pushed the fix_allreduce_fusion_cp branch from aba32e3 to 99089e9 Compare April 8, 2026 08:10
@Shunkangz
Copy link
Copy Markdown
Contributor Author

/tag-and-rerun-ci

@Shunkangz
Copy link
Copy Markdown
Contributor Author

/rerun-failed-ci

@Fridge003
Copy link
Copy Markdown
Collaborator

@Shunkangz
Copy link
Copy Markdown
Contributor Author

/rerun-failed-ci

@Shunkangz
Copy link
Copy Markdown
Contributor Author

@Shunkangz Please check this failure https://github.com/sgl-project/sglang/actions/runs/24327366445/job/71025454116?pr=21249

It's a merge artifact. I fixed it.

@Shunkangz
Copy link
Copy Markdown
Contributor Author

/rerun-failed-ci

@Fridge003 Fridge003 force-pushed the fix_allreduce_fusion_cp branch from c684655 to 6ef8499 Compare April 16, 2026 04:46
@Fridge003 Fridge003 merged commit e389a52 into sgl-project:main Apr 20, 2026
254 of 305 checks passed
zhangying098 pushed a commit to zhangying098/sglang that referenced this pull request Apr 23, 2026
Co-authored-by: Shunkang <182541032+Shunkangz@users.noreply.github.co>
kyx1999 pushed a commit to KMSorSMS/sglang that referenced this pull request Apr 27, 2026
Co-authored-by: Shunkang <182541032+Shunkangz@users.noreply.github.co>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants