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
[doc] update `vllm serve` cli args documentation (#24329)
Signed-off-by: cjackal <44624812+cjackal@users.noreply.github.com>
cjackal committed
Sep 9, 2025 at 12:07 UTC
13b89bd823b652979a8b2b446e1860aeb5e6a538
1 file changed
+4
-5
vllm/entrypoints/openai/cli_args.py
+4
-5
@@ -134,14 +134,13 @@ schema. Example: `[{"type": "text", "text": "Hello world!"}]`"""
134
"""If specified, will run the OpenAI frontend server in the same process as
135
the model serving engine."""
136
enable_request_id_headers: bool = False
137
- """If specified, API server will add X-Request-Id header to responses.
138
- Caution: this hurts performance at high QPS."""
137
+ """If specified, API server will add X-Request-Id header to responses."""
138
enable_auto_tool_choice: bool = False
140
- """If specified, exclude tool definitions in prompts when
141
- tool_choice='none'."""
142
- exclude_tools_when_tool_choice_none: bool = False
139
"""Enable auto tool choice for supported models. Use `--tool-call-parser`
140
to specify which parser to use."""
141
+ exclude_tools_when_tool_choice_none: bool = False
142
+ """If specified, exclude tool definitions in prompts when
143
+ tool_choice='none'."""
144
tool_call_parser: Optional[str] = None
145
"""Select the tool call parser depending on the model that you're using.
146
This is used to parse the model-generated tool call into OpenAI API format.