Zetong Li 179b897b52 [Bugfix][MTP] Fix performance degradation when mtp>1 (#3597) ...

### What this PR does / why we need it?
This PR aims to fix performance degradation when mtp>1. Since mtp>1 may
result in more tokens (i.e. larger batch size) than acl graph maximum
batch size, this will cause draft model to run in eager mode.

- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0

---------

Signed-off-by: Zetong Li <slippersss@126.com>

2025-10-22 22:04:43 +08:00

..

e2e

[CI] Multi-Node CI scalable (#3611)

2025-10-22 14:18:43 +08:00

ut

[Bugfix][MTP] Fix performance degradation when mtp>1 (#3597)

2025-10-22 22:04:43 +08:00

__init__.py

[SpecDecode] Add spec decode support (#500)

2025-04-17 20:16:32 +08:00