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 15, 21:00 UTC
[Bugfix] Restrict MacOS CPU detection (#14210)
Signed-off-by: mgoin <mgoin64@gmail.com>
Michael Goin committed
Mar 4, 2025 at 09:25 UTC
6247bae6c69de73e2c8a9964d23357bef724ee16
1 file changed
+3
-3
vllm/platforms/__init__.py
+3
-3
@@ -160,11 +160,11 @@ def cpu_platform_plugin() -> Optional[str]:
160
logger.debug("Confirmed CPU platform is available because"
161
" vLLM is built with CPU.")
162
if not is_cpu:
163
- import platform
164
- is_cpu = platform.machine().lower().startswith("arm")
163
+ import sys
164
+ is_cpu = sys.platform.startswith("darwin")
165
if is_cpu:
166
logger.debug("Confirmed CPU platform is available"
167
- " because the machine is ARM.")
167
+ " because the machine is MacOS.")
168
169
except Exception as e:
170
logger.debug("CPU platform is not available because: %s", str(e))