Skip to content

Conversation

@whx-sjtu
Copy link
Collaborator

@whx-sjtu whx-sjtu commented Oct 30, 2025

What this PR does / why we need it?

This PR fixes deepseek v3.2 mtp bug.

Does this PR introduce any user-facing change?

None

How was this patch tested?

All existed ci tests should pass.

@github-actions
Copy link

👋 Hi! Thank you for contributing to the vLLM Ascend project. The following points will speed up your PR merge:‌‌

  • A PR should do only one thing, smaller PRs enable faster reviews.
  • Every PR should include unit tests and end-to-end tests ‌to ensure it works and is not broken by other future PRs.
  • Write the commit message by fulfilling the PR description to help reviewer and future developers understand.

If CI fails, you can run linting and testing checks locally according Contributing and Testing.

Copy link
Contributor

@gemini-code-assist gemini-code-assist bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Code Review

This PR addresses a bug in the deepseek v3.2 mtp implementation by incorporating the DeepseekV32IndexerCache class and adjusting the logic for identifying attention layers in the draft model. The changes ensure that the indexer layers are correctly excluded from the draft attention layers, preventing potential errors during speculative decoding.

Signed-off-by: whx-sjtu <2952154980@qq.com>
@whx-sjtu whx-sjtu added ready read for review ready-for-test start test by label for PR labels Oct 31, 2025
Signed-off-by: whx-sjtu <2952154980@qq.com>
@wangxiyuan wangxiyuan merged commit e9bb449 into vllm-project:main Nov 4, 2025
24 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

ready read for review ready-for-test start test by label for PR

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants