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
[Misc]fixed disable these http request logs (#14754)
Signed-off-by: chaunceyjiang <chaunceyjiang@gmail.com>
Chauncey committed
Mar 20, 2025 at 12:53 UTC
ae65f3e237a3a2690335e9fc12bde69cbe205062
2 files changed
+6
vllm/entrypoints/openai/api_server.py
+3
@@ -1036,6 +1036,9 @@ async def run_server(args, **uvicorn_kwargs) -> None:
1036
host=args.host,
1037
port=args.port,
1038
log_level=args.uvicorn_log_level,
1039
+ # NOTE: When the 'disable_uvicorn_access_log' value is True,
1040
+ # no access log will be output.
1041
+ access_log=not args.disable_uvicorn_access_log,
1042
timeout_keep_alive=TIMEOUT_KEEP_ALIVE,
1043
ssl_keyfile=args.ssl_keyfile,
1044
ssl_certfile=args.ssl_certfile,
vllm/entrypoints/openai/cli_args.py
+3
@@ -89,6 +89,9 @@ def make_arg_parser(parser: FlexibleArgumentParser) -> FlexibleArgumentParser:
89
default="info",
90
choices=['debug', 'info', 'warning', 'error', 'critical', 'trace'],
91
help="Log level for uvicorn.")
92
+ parser.add_argument("--disable-uvicorn-access-log",
93
+ action="store_true",
94
+ help="Disable uvicorn access log.")
95
parser.add_argument("--allow-credentials",
96
action="store_true",
97
help="Allow credentials.")