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, 19:30 UTC
[ BugFix ] Fix ZMQ when `VLLM_PORT` is set (#7205)
Robert Shaw committed
Aug 6, 2024 at 12:26 UTC
541c1852d37b9502fbc06253def70e901ca0c352
1 file changed
+1
-1
vllm/envs.py
+1
-1
@@ -145,7 +145,7 @@ environment_variables: Dict[str, Callable[[], Any]] = {
145
# used when the frontend api server is running in multi-processing mode,
146
# to communicate with the backend engine process over ZMQ.
147
'VLLM_RPC_PORT':
148
- lambda: int(os.getenv('VLLM_PORT', '5570')),
148
+ lambda: int(os.getenv('VLLM_RPC_PORT', '5570')),
149
150
# If true, will load models from ModelScope instead of Hugging Face Hub.
151
# note that the value is true or false, not numbers