Init attention backend for Intel XPU (#10656)

Co-authored-by: guangyey <guangye.yu@intel.com>
Co-authored-by: DiweiSun <105627594+DiweiSun@users.noreply.github.com>
This commit is contained in:
Meng, Hengyu
2025-10-21 11:41:28 +08:00
committed by GitHub
parent fb6cc7b000
commit b113c72e7a
18 changed files with 1210 additions and 26 deletions

View File

@@ -1,5 +1,3 @@
# xpu is not enabled in public vllm and torch whl,
# need to follow https://docs.vllm.ai/en/latest/getting_started/xpu-installation.html install vllm
[build-system]
requires = ["setuptools>=61.0", "wheel"]
build-backend = "setuptools.build_meta"
@@ -17,6 +15,10 @@ classifiers = [
]
dependencies = [
"torch==2.8.0",
"torchaudio==2.8.0",
"torchvision",
"sgl-kernel @ git+https://github.com/sgl-project/sgl-kernel-xpu.git",
"IPython",
"aiohttp",
"anthropic>=0.20.0",
@@ -61,7 +63,7 @@ dependencies = [
"transformers==4.57.1",
"uvicorn",
"uvloop",
"xgrammar==0.1.25",
# "xgrammar==0.1.24", , xgrammar depends on CUDA PyTorch and Triton only
"grpcio==1.75.1", # keep it align with compile_proto.py
"grpcio-tools==1.75.1", # keep it align with compile_proto.py
"grpcio-reflection==1.75.1", # required by srt/entrypoints/grpc_server.py