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][V1][Minor] Fix shutting_down flag checking in V1 MultiprocExecutor (#14053)
Li, Jiang committed
Mar 1, 2025 at 14:31 UTC
02296f420d38fdc5a8d9345679ece47f31c85ad2
1 file changed
+1
-1
vllm/v1/executor/multiproc_executor.py
+1
-1
@@ -170,7 +170,7 @@ class MultiprocExecutor(Executor):
170
171
def shutdown(self):
172
"""Properly shut down the executor and its workers"""
173
- if getattr(self, 'shutting_down', False):
173
+ if not getattr(self, 'shutting_down', False):
174
self.shutting_down = True
175
for w in self.workers:
176
w.worker_response_mq = None