Method __call__

llama_cpp/llama_chat_format.py:258–300

(
        self,
        *,
        messages: List[llama_types.ChatCompletionRequestMessage],
        functions: Optional[List[llama_types.ChatCompletionFunction]] = None,
        function_call: Optional[llama_types.ChatCompletionRequestFunctionCall] = None,
        tools: Optional[List[llama_types.ChatCompletionTool]] = None,
        tool_choice: Optional[llama_types.ChatCompletionToolChoiceOption] = None,
        **kwargs: Any,
    )

        )

    def __call__(
        self,
        *,
        messages: List[llama_types.ChatCompletionRequestMessage],
        functions: Optional[List[llama_types.ChatCompletionFunction]] = None,
        function_call: Optional[llama_types.ChatCompletionRequestFunctionCall] = None,
        tools: Optional[List[llama_types.ChatCompletionTool]] = None,
        tool_choice: Optional[llama_types.ChatCompletionToolChoiceOption] = None,
        **kwargs: Any,
    ) -> ChatFormatterResponse:
        def raise_exception(message: str):
            raise ValueError(message)

        prompt = self._environment.render(
            messages=messages,
            eos_token=self.eos_token,
            bos_token=self.bos_token,
            raise_exception=raise_exception,
            add_generation_prompt=self.add_generation_prompt,
            functions=functions,
            function_call=function_call,
            tools=tools,
            tool_choice=tool_choice,
            strftime_now=self.strftime_now,
            **kwargs,
        )

        stopping_criteria = None
        if self.stop_token_ids is not None:

            def stop_on_last_token(
                tokens: npt.NDArray[np.intc], logits: npt.NDArray[np.single]
            ) -> bool:
                return tokens[-1] in self.stop_token_ids

            stopping_criteria = llama.StoppingCriteriaList([stop_on_last_token])

        return ChatFormatterResponse(
            prompt=prompt,
            stop=[self.eos_token],
            stopping_criteria=stopping_criteria,
            added_special=True,
        )

    def to_chat_handler(self) -> LlamaChatCompletionHandler:
        return chat_formatter_to_chat_completion_handler(self)