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, 18:30 UTC
[BugFix][Typing] Fix Imprecise Type Annotations (#15208)
Signed-off-by: Wang Ran (汪然) <wrran@outlook.com>
Wang Ran (汪然) committed
Mar 23, 2025 at 00:05 UTC
dd861b992f33f3f4a06ee91cf3346b038929d000
1 file changed
+3
-3
vllm/v1/engine/core_client.py
+3
-3
@@ -212,9 +212,9 @@ class BackgroundResources:
212
"""Used as a finalizer for clean shutdown, avoiding
213
circular reference back to the client object."""
214
215
- ctx: Union[zmq.Context] = None
216
- output_socket: Union[zmq.Socket, zmq.asyncio.Socket] = None
217
- input_socket: Union[zmq.Socket, zmq.asyncio.Socket] = None
215
+ ctx: zmq.Context
216
+ output_socket: Optional[Union[zmq.Socket, zmq.asyncio.Socket]] = None
217
+ input_socket: Optional[Union[zmq.Socket, zmq.asyncio.Socket]] = None
218
proc_handle: Optional[BackgroundProcHandle] = None
219
shutdown_path: Optional[str] = None
220