@tduncan / Nvidia-Optimization-Stack-VLLM Mirror · read-only / commits / ccb5376a9a
Fork 0

A high-throughput and memory-efficient inference and serving engine for LLMs

Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git

Synced Nov 16, 08:15 UTC

[Bugfix][OpenVINO] Fix circular reference #9939 (#9974)

Signed-off-by: MengqingCao <cmq0113@163.com>

Mengqing Cao committed Nov 4, 2024 at 18:14 UTC ccb5376a9a88bb6251c4434b79c173151e6f7729
1 file changed +4 -2
vllm/platforms/openvino.py
+4 -2
@@ -1,10 +1,12 @@
1 import torch
2
3 import vllm.envs as envs
4 -from vllm.utils import print_warning_once
4 +from vllm.logger import init_logger
5
6 from .interface import Platform, PlatformEnum
7
8 +logger = init_logger(__name__)
9 +
10
11 class OpenVinoPlatform(Platform):
12 _enum = PlatformEnum.OPENVINO
@@ -27,5 +29,5 @@ class OpenVinoPlatform(Platform):
29
30 @classmethod
31 def is_pin_memory_available(self) -> bool:
30 - print_warning_once("Pin memory is not supported on OpenViNO.")
32 + logger.warning("Pin memory is not supported on OpenViNO.")
33 return False