Skip to content

[Bugfix] Default Qwen3 reasoning parser to prompt-has-open-think #52

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
rivetphilbot wants to merge 1 commit into from
Closed

[Bugfix] Default Qwen3 reasoning parser to prompt-has-open-think #52

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 5 additions & 3 deletions vllm/reasoning/qwen3_reasoning_parser.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -27,9 +27,11 @@ class Qwen3ReasoningParser(BaseThinkingReasoningParser):
def __init__(self, tokenizer, *args, **kwargs):
super().__init__(tokenizer, *args, **kwargs)
chat_kwargs = kwargs.get("chat_template_kwargs", {}) or {}
# Qwen3.5 chat templates open the <think> block in the prompt when
# thinking is enabled, so completion tokens may only contain </think>.
self.prompt_has_open_think = bool(chat_kwargs.get("enable_thinking", False))
# The standard Qwen3 chat template injects <think>\n into the prompt
# whenever enable_thinking is not explicitly False, so completion tokens
# may only contain </think>. Default to True and only honor an explicit
# enable_thinking=False to opt out.
self.prompt_has_open_think = bool(chat_kwargs.get("enable_thinking", True))

@property
def start_token(self) -> str:
Expand Down
Loading