Cherry pick weight_checker _weight_fp32 buffer skip from #22663 - #24534
Merged
Merged
Conversation
Contributor
|
Warning You have reached your daily quota limit. Please wait up to 24 hours and I will start processing your requests again! |
(cherry picked from commit 98b5440)
Collaborator
Author
|
/tag-and-rerun-ci |
Cherry-picked weight_checker.py changes only from the original PR (which also touched many other files). (cherry picked from commit 629aa25)
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com> Co-authored-by: Letian Ruan <ruanlt420@gmail.com> Cherry-picked weight_checker.py changes only from the original PR. (cherry picked from commit 2b93c2f)
fzyzcjy
force-pushed
the
cherry_pick_weight_checker/2
branch
from
May 6, 2026 13:09
2fa6c33 to
37b6e23
Compare
fzyzcjy
changed the title
_weight_fp32 buffer skip from #22663
LLThomas
pushed a commit
to LLThomas/sglang
that referenced
this pull request
May 8, 2026
…t#22663 (sgl-project#24534) Co-authored-by: JD <jaedon.guo@gmail.com>
Chronostasys
pushed a commit
to MindLab-Research/sglang
that referenced
this pull request
Aug 24, 2026
…t#22663 (sgl-project#24534) Co-authored-by: JD <jaedon.guo@gmail.com>
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
Cherry-picks only the
weight_checker.pyportion of #22663 ("[Hotfix] final fixes for P2P Transfer") fromsglang-milestomain. The original PR also touched several unrelated files (model_runner, deepseek_weight_loader, qwen3); those are intentionally excluded here.Original author: @JDGUO (JD). I'm only the courier.
Changes
Adds
"_weight_fp32"to both:_reset_tensors(alongsidecos_sin_cache/freqs_cis), andnon_persistent_buffer_patternslist in_postprocess_tensors._weight_fp32is the FP32 cache of the gate weight (e.g.Glm4MoeGate), which is non-persistent and recomputed.PR chain
3rd of 3. Stacked on #24533, which is stacked on #24532. After both prior PRs land, this PR's diff will collapse to only this PR's
weight_checker.pychange.Test plan