EngineX/xc-llm-ascend
3
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

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

Browse Source 
### What this PR does / why we need it?
Fix the bug " TypeError: 'NoneType' object is not iterable' " in
vllm_ascend/compilation/acl_graph.py
The reason of that is the attn_metadata is none in the dummy_run of MTP.

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

Signed-off-by: chenmenglong <chenmenglong1@huawei.com>
This commit is contained in:
MengLong Chen
2025-12-16 09:14:05 +08:00
committed by GitHub
parent d43cabc2b1
commit 5e0ada5395
6 changed files with 8 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
vllm_ascend/worker/model_runner_v1.py
View File

@@ -2296,7 +2296,7 @@ class NPUModelRunner(GPUModelRunner):
aclgraph_runtime_mode=aclgraph_runtime_mode,
batch_descriptor=batch_descriptor,
dummy_compute_logits=dummy_drafter_compute_logits,
skip_attn=not force_attention)
in_graph_capturing=not force_attention)
if self.in_profile_run and self.dynamic_eplb:
self.model.clear_all_moe_loads()
if not self.in_profile_run and self.dynamic_eplb: