vllm.entrypoints.openai.utils ¶
_ChatCompletionResponseChoiceT module-attribute ¶
_ChatCompletionResponseChoiceT = TypeVar(
"_ChatCompletionResponseChoiceT",
ChatCompletionResponseChoice,
ChatCompletionResponseStreamChoice,
)
maybe_filter_parallel_tool_calls ¶
maybe_filter_parallel_tool_calls(
choice: _ChatCompletionResponseChoiceT,
request: ChatCompletionRequest,
) -> _ChatCompletionResponseChoiceT
Filter to first tool call only when parallel_tool_calls is False.