Skip to content

support PCP #25852

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.

Sign up for GitHub

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

Jump to bottom
Closed
wants to merge 31 commits into from
Closed

support PCP #25852

wants to merge 31 commits into from

Conversation

@LookAround0301
Copy link

@LookAround0301 LookAround0301 commented Sep 29, 2025 •
edited by github-actions bot
Loading

Purpose

Test Plan

Test Result

Essential Elements of an Effective PR Description Checklist
  • The purpose of the PR, such as "Fix some issue (link existing issues this PR will resolve)".
  • The test plan, such as providing test command.
  • The test results, such as pasting the results comparison before and after, or e2e results
  • (Optional) The necessary documentation update, such as updating supported_models.md and examples for a new model.
  • (Optional) Release notes update. If your change is user facing, please update the release notes draft in the Google Doc.
@LookAround0301
[scheduler] support cp
6a84571 
Signed-off-by: LookAround <lixushi@huawei.com>
@LookAround0301
[bug fix] fix block size bug
4ec01d7 
Signed-off-by: LookAround <lixushi@huawei.com>
@mergify mergify bot added the v1 label Sep 29, 2025
LookAround0301 and others added 13 commits September 29, 2025 21:27
@LookAround0301
[bug fix] fix block size bug
ccb6239 
Signed-off-by: LookAround <lixushi@huawei.com>
@FENP
flashinfer backend support context parallel — decode phase
f064d1a
@FENP
flashinfer backend support context parallel — prefill phase
f649ffd
@pisceskkk @FENP
flashinfer backend support context parallel — custom mask and kv reorg
d604ae3
@FENP
add context_parallel_size to ensure_model_parallel_initialized
c0bbb52
@FENP
bug fix: only update decode seq_lens
3f418de
@pisceskkk @FENP
bug fix: update custom mask calc
817bc28
@LookAround0301 @FENP
[scheduler] support cp
3add9a3 
Signed-off-by: LookAround <lixushi@huawei.com>
@LookAround0301 @FENP
[bug fix] fix block size bug
a85aa69 
Signed-off-by: LookAround <lixushi@huawei.com>
@LookAround0301 @FENP
[bug fix] fix block size bug
880ddbf 
Signed-off-by: LookAround <lixushi@huawei.com>
@pisceskkk
support cp for flashinfer
de86d38 
Signed-off-by: QiuChunshuo <qiuchunshuo@huawei.com>
@pisceskkk
Merge remote-tracking branch 'origin/context-parallel-flashinfer' int…
abc6470 
…o long_seq_dev Signed-off-by: QiuChunshuo <qiuchunshuo@huawei.com>
@LookAround0301
[scheduler] support cp
f8afd97 
Signed-off-by: LookAround <lixushi@huawei.com>
@mergify
Copy link

mergify bot commented Oct 3, 2025

This pull request has merge conflicts that must be resolved before it can be
merged. Please rebase the PR, @LookAround0301.

https://docs.github.com/en/pull-requests/collaborating-with-pull-requests/working-with-forks/syncing-a-fork
@mergify mergify bot added the needs-rebase label Oct 3, 2025
@pisceskkk
disable cudagraph global for cp
901028a 
Signed-off-by: QiuChunshuo <qiuchunshuo@huawei.com>
@pisceskkk
[Feature] support multi-requests
cf4d3ab 
Signed-off-by: QiuChunshuo <qiuchunshuo@huawei.com>
@pisceskkk
[typo] remove unused parameters
07f9920 
Signed-off-by: QiuChunshuo <qiuchunshuo@huawei.com>
@pisceskkk
[ModelRunner] Refactor _update_tokens_for_cp
1407678 
Signed-off-by: QiuChunshuo <qiuchunshuo@huawei.com>
@pisceskkk
[bugfix] type convert error
b40f23e 
Signed-off-by: QiuChunshuo <qiuchunshuo@huawei.com>
@zhangsicheng5 zhangsicheng5 mentioned this pull request Oct 14, 2025
Merged
5 tasks
@pisceskkk pisceskkk mentioned this pull request Oct 14, 2025
Open
9 tasks
pisceskkk and others added 8 commits October 14, 2025 13:06
@pisceskkk
[flashinfer] Vectorized custom_mask computation
cdf2495 
Signed-off-by: QiuChunshuo <qiuchunshuo@huawei.com>
@pisceskkk
[bugfix] add cp_world_size&cp_rank to AttentionImpl
50ab73c 
Signed-off-by: QiuChunshuo <qiuchunshuo@huawei.com>
@pisceskkk
[format] break too long lines and add assert for variables of uncerta…
67b052c 
…in type Signed-off-by: QiuChunshuo <qiuchunshuo@huawei.com>
@pisceskkk
[format] Rename to distinguish different types of variables
b461a55 
Signed-off-by: QiuChunshuo <qiuchunshuo@huawei.com>
@pisceskkk
[format] add error for unreachable branch.
c9fd35a 
Signed-off-by: QiuChunshuo <qiuchunshuo@huawei.com>
@pisceskkk
[Perf] Persist some variables that require round-by-round initializat…
3212997 
…ion. Signed-off-by: QiuChunshuo <qiuchunshuo@huawei.com>
@pisceskkk
[format] isort
51cb1f2 
Signed-off-by: QiuChunshuo <qiuchunshuo@huawei.com>
@LookAround0301
Merge pull request #3 from pisceskkk/long_seq_dev
9f8290e 
[Feature] support multi-requests
@mergify mergify bot mentioned this pull request Oct 15, 2025
Open
3 tasks
@FENP
Copy link
Contributor

FENP commented Oct 15, 2025

A new open PR #26864 replaces this PR.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

needs-rebase v1

3 participants

@LookAround0301 @FENP @pisceskkk