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, 22:15 UTC
[MISC] Logging the message about Ray teardown (#13502)
Signed-off-by: Cody Yu <hao.yu.cody@gmail.com> Co-authored-by: Rui Qiao <161574667+ruisearch42@users.noreply.github.com>
Cody Yu committed
Feb 19, 2025 at 09:40 UTC
fbbe1fbac669a17f81c74c696f11a51167ed6a3c
1 file changed
+4
vllm/executor/ray_distributed_executor.py
+4
@@ -101,6 +101,10 @@ class RayDistributedExecutor(DistributedExecutorBase):
101
self.driver_worker.execute_method)
102
103
def shutdown(self) -> None:
104
+ logger.info(
105
+ "Shutting down Ray distributed executor. If you see error log "
106
+ "from logging.cc regarding SIGTERM received, please ignore because "
107
+ "this is the expected termination process in Ray.")
108
if hasattr(self, "forward_dag") and self.forward_dag is not None:
109
self.forward_dag.teardown()
110
import ray