Wang Yixuan 153eeaa621 [Bugfix] Fix DeepSeek FIA error in async_scheduling with mtp (#5046) ...

### What this PR does / why we need it?
When enable the async_scheduling, in large scale EP scene, mtp module
goes to eagler mode, which results in the mismatch of
seq_lens_list、block_table. So adapt the judgement before the draft model
forward.

fix #4986 

### Does this PR introduce _any_ user-facing change?
No

### How was this patch tested?

- vLLM version: v0.12.0
- vLLM main:
ad32e3e19c

Signed-off-by: hust17yixuan <303660421@qq.com>

2025-12-17 09:20:44 +08:00

..

__init__.py

Drop torchair (#4814)

2025-12-10 09:20:40 +08:00

eagle_proposer.py

[Fusion] [Graph] Add qknorm rope fusion operator (#4711)

2025-12-17 08:53:44 +08:00

interface.py

[Bugfix] Fix the attn_metadata is None (#5038)

2025-12-16 09:14:05 +08:00

mtp_proposer.py

[Bugfix] Fix DeepSeek FIA error in async_scheduling with mtp (#5046)

2025-12-17 09:20:44 +08:00

ngram_proposer.py

[Bugfix] Fix the attn_metadata is None (#5038)

2025-12-16 09:14:05 +08:00

suffix_proposer.py

[Bugfix] Fix the attn_metadata is None (#5038)

2025-12-16 09:14:05 +08:00