xpu_model_runner.py

# SPDX-License-Identifier: Apache-2.0
from typing import TYPE_CHECKING

import torch

from vllm.config import VllmConfig
from vllm.logger import init_logger
from vllm.v1.worker.gpu_model_runner import GPUModelRunner

if TYPE_CHECKING:
    pass

logger = init_logger(__name__)


class XPUModelRunner(GPUModelRunner):
    """A model runner for XPU devices."""

    def __init__(
        self,
        vllm_config: VllmConfig,
        device: torch.device,
    ):
        super().__init__(vllm_config, device)
        # FIXME: To be verified.
        self.cascade_attn_enabled = False

    def _init_device_properties(self) -> None:
        pass

    def _sync_device(self) -> None:
        torch.xpu.synchronize()