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 14, 05:30 UTC
[Bugfix][Frontend] Fix `--enable-log-outputs` does not match the documentation (#24626)
Signed-off-by: Kebe <mail@kebe7jun.com>
Kebe committed
Sep 13, 2025 at 00:01 UTC
684b6870e1af43257ecac2bd0ed869bae163c89a
1 file changed
+5
-2
vllm/entrypoints/openai/cli_args.py
+5
-2
@@ -171,8 +171,8 @@ schema. Example: `[{"type": "text", "text": "Hello world!"}]`"""
171
"""Enable the /get_tokenizer_info endpoint. May expose chat
172
templates and other tokenizer configuration."""
173
enable_log_outputs: bool = False
174
- """If set to True, enable logging of model outputs (generations)
175
- in addition to the input logging that is enabled by default."""
174
+ """If True, log model outputs (generations).
175
+ Requires --enable-log-requests."""
176
h11_max_incomplete_event_size: int = H11_MAX_INCOMPLETE_EVENT_SIZE_DEFAULT
177
"""Maximum size (bytes) of an incomplete HTTP event (header or body) for
178
h11 parser. Helps mitigate header abuse. Default: 4194304 (4 MB)."""
@@ -273,6 +273,9 @@ def validate_parsed_serve_args(args: argparse.Namespace):
273
if args.enable_auto_tool_choice and not args.tool_call_parser:
274
raise TypeError("Error: --enable-auto-tool-choice requires "
275
"--tool-call-parser")
276
+ if args.enable_log_outputs and not args.enable_log_requests:
277
+ raise TypeError("Error: --enable-log-outputs requires "
278
+ "--enable-log-requests")
279
280
281
def create_parser_for_docs() -> FlexibleArgumentParser: