@tduncan / Nvidia-Optimization-Stack-VLLM Mirror · read-only / commits / 47826cacf0
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, 07:00 UTC

[Bugfix] Ignore ray reinit error when current platform is ROCm or XPU (#10375)

Signed-off-by: Hollow Man <hollowman@opensuse.org>

ℍ𝕠𝕝𝕝𝕠𝕨 𝕄𝕒𝕟 committed Nov 18, 2024 at 05:29 UTC 47826cacf0e037b4e109f0b2d8d594e47def500e
1 file changed +1 -1
vllm/executor/ray_utils.py
+1 -1
@@ -234,7 +234,7 @@ def initialize_ray_cluster(
234 if current_platform.is_rocm() or current_platform.is_xpu():
235 # Try to connect existing ray instance and create a new one if not found
236 try:
237 - ray.init("auto")
237 + ray.init("auto", ignore_reinit_error=True)
238 except ConnectionError:
239 logger.warning(
240 "No existing RAY instance detected. "