biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Frontend] Enforce tokenize=False when applying chat template (#27205)

Browse Source 
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
Co-authored-by: Isotr0py <mozf@mail2.sysu.edu.cn>
This commit is contained in:
Russell Bryant
2025-10-20 22:57:34 -04:00
committed by GitHub
parent 0eb8f2b880
commit 3ada34f9cb
2 changed files with 31 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files

31
vllm/entrypoints/chat_utils.py
View File

@@ -1499,18 +1499,25 @@ def resolve_chat_template_kwargs(
tokenizer: PreTrainedTokenizer | PreTrainedTokenizerFast,
chat_template: str,
chat_template_kwargs: dict[str, Any],
raise_on_unexpected: bool = True,
) -> dict[str, Any]:
# We exclude chat_template from kwargs here, because
# chat template has been already resolved at this stage
unexpected_vars = {"chat_template", "tokenize"}
if raise_on_unexpected and (
unexpected_in_kwargs := unexpected_vars & chat_template_kwargs.keys()
):
raise ValueError(
"Found unexpected chat template kwargs from request: "
f"{unexpected_in_kwargs}"
)
fn_kw = {
k
for k in chat_template_kwargs
if supports_kw(tokenizer.apply_chat_template, k, allow_var_kwargs=False)
}
template_vars = _cached_resolve_chat_template_kwargs(chat_template)
# We exclude chat_template from kwargs here, because
# chat template has been already resolved at this stage
unexpected_vars = {"chat_template"}
accept_vars = (fn_kw | template_vars) - unexpected_vars
return {k: v for k, v in chat_template_kwargs.items() if k in accept_vars}
@@ -1522,7 +1529,6 @@ def apply_hf_chat_template(
tools: list[dict[str, Any]] | None,
*,
model_config: ModelConfig,
tokenize: bool = False, # Different from HF's default
**kwargs: Any,
) -> str:
hf_chat_template = resolve_hf_chat_template(
@@ -1539,17 +1545,18 @@ def apply_hf_chat_template(
"does not define one."
)
resolved_kwargs = resolve_chat_template_kwargs(
tokenizer=tokenizer,
chat_template=hf_chat_template,
chat_template_kwargs=kwargs,
)
try:
resolved_kwargs = resolve_chat_template_kwargs(
tokenizer=tokenizer,
chat_template=hf_chat_template,
chat_template_kwargs=kwargs,
)
return tokenizer.apply_chat_template(
conversation=conversation, # type: ignore[arg-type]
tools=tools, # type: ignore[arg-type]
chat_template=hf_chat_template,
tokenize=tokenize,
tokenize=False,
**resolved_kwargs,
)