xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2025-12-10 04:44:57 +08:00
Code Issues Packages Projects Releases Wiki Activity

[Intel GPU] support ray as distributed executor backend for XPU. (#20659)

Browse Source 
Signed-off-by: Kunshang Ji <kunshang.ji@intel.com>
This commit is contained in:
Kunshang Ji 2025-07-09 15:36:58 +08:00 committed by GitHub
parent e760fcef22
commit b6e7e3d58f
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
2 changed files with 3 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
.buildkite/scripts/hardware_ci/run-xpu-test.sh
View File

@ -27,6 +27,8 @@ docker run \
"${image_name}" \
sh -c '
VLLM_USE_V1=1 python3 examples/offline_inference/basic/generate.py --model facebook/opt-125m --block-size 64 --enforce-eager
VLLM_USE_V1=1 python3 examples/offline_inference/basic/generate.py --model facebook/opt-125m --block-size 64 --enforce-eager -tp 2 --distributed-executor-backend ray
VLLM_USE_V1=1 python3 examples/offline_inference/basic/generate.py --model facebook/opt-125m --block-size 64 --enforce-eager -tp 2 --distributed-executor-backend mp
cd tests
pytest -v -s v1/core
'

2
vllm/executor/ray_distributed_executor.py
View File

@ -62,7 +62,7 @@ class RayDistributedExecutor(DistributedExecutorBase):
def _init_executor(self) -> None:
self.forward_dag: Optional[ray.dag.CompiledDAG] = None
if envs.VLLM_USE_V1 and not current_platform.is_xpu():
if envs.VLLM_USE_V1:
# V1 uses SPMD worker and compiled DAG
os.environ["VLLM_USE_RAY_SPMD_WORKER"] = "1"
os.environ["VLLM_USE_RAY_COMPILED_DAG"] = "1"