Repository navigation
refactor: inject scheduler and EPLB options across runtime boundaries. - #2738
Open
liutongxuan wants to merge 3 commits into
Open
liutongxuan wants to merge 3 commits into
liutongxuan wants to merge 3 commits into
Conversation
liutongxuan
requested review from
DongheJin,
DragonFive,
JimHsiung,
Kang-Meng,
XuZhang99,
xiao-yu-chen,
yingxudeng and
zhang-minchao
as code owners
November 13, 2026 21:35
Codex Review Summary: ❌ FailedThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
liutongxuan
changed the title
Serialize producer-owned expert loads and prepare tokens without consulting process-wide enablement. Keep tensor validation and shared-memory capacity checks at the transport boundary. Cover LLM and VLM payloads, token-only results, empty metadata, and conflicting global settings in both transports.
Project mix-batch enablement and effective KV split into SchedulerOptions at the LLM and VLM master boundaries. Batch-mode selection then depends on its input while preserving chunking, parallel-layout, speculative-decoding, and PD role rules. Exercise conflicting process settings and migrate existing mode-specific tests to explicit inputs without changing chunk-size resolution.
Resolve worker enablement, redundant expert count, and decode-only measurement once at the LLM master. Carry this snapshot through in-process and spawned worker bootstrap, collective communicators, and model context into native MoE layers and the Python bridge. Preserve load buffers for both target and draft workers while keeping the EPLB executor target-only. Keep expert parallel degree and rank-table configuration in their existing communication bootstrap paths. Validate appended spawn arguments at the process boundary and retain legacy defaults. Cover conflicting globals, snapshot copies, native expert storage, Python configuration, overlap metadata, and target/draft initialization.
liutongxuan
force-pushed
the
features/refactor_options3
branch
from
November 15, 2026 08:04
79a6b8c to
9cc6e6e
Compare
This branch has not been deployed
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
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.
Description
Related Issues
Change Type
Pull Request Checklist
Thank you for contributing to xLLM. Before requesting review, please make sure the following items are complete.
PR Title and Commit Messages
<type>: <subject>.Pre-commit Checks
pre-commitby runningpip install pre-commitor an equivalent command.pre-commit install.pre-commit run --all-filesand fixed any reported issues.Self Review
.agents/skills/code-review/references/custom-code-style.md, especially code written or assisted by AI.mainbranch.Build and Test Coverage
python setup.py build testhas passed on a CUDA machine.python setup.py build testhas passed on an NPU machine.python setup.py build testhas passed on an MLU machine.Reviewer Notes