xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2025-12-27 16:35:17 +08:00
Code Issues Packages Projects Releases Wiki Activity

[Misc][XPU] Avoid torch compile for XPU platform (#10747)

Browse Source 
Signed-off-by: yan ma <yan.ma@intel.com>
Co-authored-by: youkaichao <youkaichao@gmail.com>
This commit is contained in:
Yan Ma 2024-12-03 01:53:55 +08:00 committed by GitHub
parent b45f0d7946
commit 519cc6ca12
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
2 changed files with 8 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
.buildkite/run-xpu-test.sh
View File

@ -12,5 +12,7 @@ remove_docker_container() { docker rm -f xpu-test || true; }
trap remove_docker_container EXIT
remove_docker_container
# Run the image and launch offline inference
docker run --network host --name xpu-test --device /dev/dri -v /dev/dri/by-path:/dev/dri/by-path --entrypoint="" xpu-test python3 examples/offline_inference.py
# Run the image and test offline inference/tensor parallel
docker run -it -d --name xpu-test --device /dev/dri -v /dev/dri/by-path:/dev/dri/by-path xpu-test /bin/bash
docker exec xpu-test bash -c "python3 examples/offline_inference.py"
docker exec xpu-test bash -c "python3 examples/offline_inference_cli.py -tp 2"

4
vllm/plugins/__init__.py
View File

@ -4,6 +4,7 @@ import os
import torch
import vllm.envs as envs
from vllm.platforms import current_platform
logger = logging.getLogger(__name__)
@ -25,6 +26,9 @@ def load_general_plugins():
os.environ['TORCHINDUCTOR_COMPILE_THREADS'] = '1'
# see https://github.com/vllm-project/vllm/issues/10619
torch._inductor.config.compile_threads = 1
if current_platform.is_xpu():
# see https://github.com/pytorch/pytorch/blob/8cada5cbe5450e17c26fb8b358116785324537b2/torch/_dynamo/config.py#L158 # noqa
os.environ['TORCH_COMPILE_DISABLE'] = 'True'
global plugins_loaded
if plugins_loaded:
return