[Rust Frontend] Correct --reasoning-parser semantics (#46359)

Signed-off-by: Bugen Zhao <i@bugenzhao.com>
This commit is contained in:
Bugen Zhao
2026-06-23 15:38:39 +08:00
committed by GitHub
parent a46f3eb232
commit 25bc3be49c
7 changed files with 232 additions and 43 deletions
+10
View File
@@ -43,6 +43,11 @@ pub struct ManagedEngineArgs {
/// Arguments after an explicit `--` are forwarded verbatim. Before `--`,
/// `vllm-rs serve` automatically keeps recognized frontend options on
/// the Rust side and forwards everything else to Python.
///
/// The explicit `--` passthrough is a last-resort escape hatch. Rust does
/// not interpret, validate, or de-duplicate those arguments against
/// managed-engine arguments that it appends later; if the same Python flag
/// appears more than once, Python argparse owns the final result.
#[arg(
last = true,
allow_hyphen_values = true,
@@ -72,6 +77,7 @@ impl ManagedEngineArgs {
model: String,
max_model_len: Option<u32>,
max_logprobs: Option<i32>,
reasoning_parser: Option<&str>,
language_model_only: bool,
disable_log_stats: bool,
shutdown_timeout: u64,
@@ -87,6 +93,10 @@ impl ManagedEngineArgs {
python_args.push("--max-logprobs".to_string());
python_args.push(max_logprobs.to_string());
}
if let Some(reasoning_parser) = reasoning_parser {
python_args.push("--reasoning-parser".to_string());
python_args.push(reasoning_parser.to_string());
}
if language_model_only {
python_args.push("--language-model-only".to_string());
}