[Hardware][Intel GPU] Add Intel GPU(XPU) inference backend (#3814)

Co-authored-by: Jiang Li <jiang1.li@intel.com>
Co-authored-by: Abhilash Majumder <abhilash.majumder@intel.com>
Co-authored-by: Abhilash Majumder <30946547+abhilash1910@users.noreply.github.com>
This commit is contained in:
Kunshang Ji
2024-06-17 11:01:25 -07:00
committed by GitHub
co-authored by Jiang Li Abhilash Majumder Abhilash Majumder
parent 1f12122b17
commit 728c4c8a06
31 changed files with 1998 additions and 24 deletions
+3 -1
View File
@@ -12,7 +12,7 @@ from vllm.model_executor.layers.quantization import QUANTIZATION_METHODS
from vllm.model_executor.models import ModelRegistry
from vllm.transformers_utils.config import get_config, get_hf_text_config
from vllm.utils import (cuda_device_count_stateless, get_cpu_memory, is_cpu,
is_hip, is_neuron, is_tpu)
is_hip, is_neuron, is_tpu, is_xpu)
if TYPE_CHECKING:
from ray.util.placement_group import PlacementGroup
@@ -757,6 +757,8 @@ class DeviceConfig:
self.device_type = "tpu"
elif is_cpu():
self.device_type = "cpu"
elif is_xpu():
self.device_type = "xpu"
else:
# We don't call torch.cuda.is_available() here to
# avoid initializing CUDA before workers are forked