vllm.reasoning.step3_reasoning_parser ¶
Step3ReasoningParser ¶
Bases: ReasoningParser
Reasoning parser for Step3 model.
The Step3 model uses token to denote the end of reasoning text. This parser extracts all content before as reasoning content.
Source code in vllm/reasoning/step3_reasoning_parser.py
__init__ ¶
__init__(
tokenizer: PreTrainedTokenizerBase, *args, **kwargs
)
Source code in vllm/reasoning/step3_reasoning_parser.py
extract_content_ids ¶
extract_reasoning_content ¶
extract_reasoning_content(
model_output: str, request: ChatCompletionRequest
) -> tuple[Optional[str], Optional[str]]
Source code in vllm/reasoning/step3_reasoning_parser.py
extract_reasoning_content_streaming ¶
extract_reasoning_content_streaming(
previous_text: str,
current_text: str,
delta_text: str,
previous_token_ids: Sequence[int],
current_token_ids: Sequence[int],
delta_token_ids: Sequence[int],
) -> Union[DeltaMessage, None]
Extract reasoning content from a delta message. Handles streaming output where previous + delta = current. Uses token IDs for faster processing. For text "abcxyz": - 'abc' goes to reasoning_content - 'xyz' goes to content