bugfix: reduce acl graph memory overhead.#1457
Open
RobbieLeung wants to merge 1 commit into
Open
Conversation
RobbieLeung
requested review from
DongheJin,
JimHsiung,
XuZhang99,
liutongxuan,
walsonyang and
yq33victor
as code owners
RobbieLeung
changed the title
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Code Review
This pull request optimizes memory allocation for persistent buffers in the ACL graph executor by sizing the attention mask based on decode graph capacity rather than the prefill budget. It also refines the capacity calculation for speculative decoding. Feedback indicates that the updated capacity function is incorrectly used for sequence-indexed metadata, causing significant memory over-allocation for tensors like block tables, which undermines the PR's memory reduction goals.
RobbieLeung
force-pushed
the
bugfix/qwen35-acl-graph-memory-overhead
branch
from
5e744c2 to
3e05202
Compare
RobbieLeung
requested review from
Clement-Wang26,
DragonFive,
Kang-Meng,
liujinguang0125,
xiao-yu-chen,
yingxudeng and
zhang-minchao
as code owners
XuZhang99
approved these changes
May 21, 2026
zhang-minchao
approved these changes
May 21, 2026
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
size acl graph persistent_mask_ by decode/spec-verify graph capacity instead of max_tokens_per_batch.