[CI] Upgrade vllm to newest commit (#3182)

### What this PR does / why we need it?
Upgrade vLLM to newest commit

- Fix the aclgraph doesn't work problem, caused by
24fab45d96
- Fix PoolerOutput import error, caused by
755ed7b05b
- Fix the aclgraph weight load error to keep the same with torchair fix.
4492e3a554

### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
All test should pass


- vLLM version: v0.10.2
- vLLM main:
52d0cb8458

---------

Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com>
This commit is contained in:
wangxiyuan
2025-09-26 06:18:15 +08:00
committed by GitHub
parent 0794f64a18
commit 2930e4a6bd
9 changed files with 49 additions and 53 deletions

View File

@@ -209,6 +209,11 @@ class NPUPlatform(Platform):
# set cudaprah sizes before extending `compilation_config.splitting_ops`
vllm_config._set_cudagraph_sizes()
# TODO: Full graph is fully supported later, and the default value will be set to full graph.
if not vllm_version_is("v0.10.2"):
if compilation_config.cudagraph_mode == CUDAGraphMode.FULL_AND_PIECEWISE:
compilation_config.cudagraph_mode = CUDAGraphMode.PIECEWISE
if compilation_config.cudagraph_mode == CUDAGraphMode.NONE:
compilation_config.level = CompilationLevel.NO_COMPILATION
# TODO: Currently MLA does not support FULL_DECODE_ONLY, remove the second condition