Method __call__

llama_cpp/llama_chat_format.py:3402–3656

(
        self,
        *,
        llama: llama.Llama,
        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,
        temperature: float = 0.2,
        top_p: float = 0.95,
        top_k: int = 40,
        min_p: float = 0.05,
        typical_p: float = 1.0,
        stream: bool = False,
        stop: Optional[Union[str, List[str]]] = [],
        seed: Optional[int] = None,
        response_format: Optional[
            llama_types.ChatCompletionRequestResponseFormat
        ] = None,
        max_tokens: Optional[int] = None,
        presence_penalty: float = 0.0,
        frequency_penalty: float = 0.0,
        repeat_penalty: float = 1.1,
        tfs_z: float = 1.0,
        mirostat_mode: int = 0,
        mirostat_tau: float = 5.0,
        mirostat_eta: float = 0.1,
        model: Optional[str] = None,
        logits_processor: Optional[llama.LogitsProcessorList] = None,
        grammar: Optional[llama.LlamaGrammar] = None,
        logit_bias: Optional[Dict[str, float]] = None,
        logprobs: Optional[bool] = None,
        top_logprobs: Optional[int] = None,
        **kwargs,  # type: ignore
    )

        return text

    def __call__(
        self,
        *,
        llama: llama.Llama,
        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,
        temperature: float = 0.2,
        top_p: float = 0.95,
        top_k: int = 40,
        min_p: float = 0.05,
        typical_p: float = 1.0,
        stream: bool = False,
        stop: Optional[Union[str, List[str]]] = [],
        seed: Optional[int] = None,
        response_format: Optional[
            llama_types.ChatCompletionRequestResponseFormat
        ] = None,
        max_tokens: Optional[int] = None,
        presence_penalty: float = 0.0,
        frequency_penalty: float = 0.0,
        repeat_penalty: float = 1.1,
        tfs_z: float = 1.0,
        mirostat_mode: int = 0,
        mirostat_tau: float = 5.0,
        mirostat_eta: float = 0.1,
        model: Optional[str] = None,
        logits_processor: Optional[llama.LogitsProcessorList] = None,
        grammar: Optional[llama.LlamaGrammar] = None,
        logit_bias: Optional[Dict[str, float]] = None,
        logprobs: Optional[bool] = None,
        top_logprobs: Optional[int] = None,
        **kwargs,  # type: ignore
    ) -> Union[
        llama_types.CreateChatCompletionResponse,
        Iterator[llama_types.CreateChatCompletionStreamResponse],
    ]:
        self._init_mtmd_context(llama)
        assert self.mtmd_ctx is not None

        image_urls = self.get_image_urls(messages)
        media_marker = self._mtmd_cpp.mtmd_default_marker().decode("utf-8")
        template_env = ImmutableSandboxedEnvironment(
            trim_blocks=True,
            lstrip_blocks=True,
            extensions=[
                Jinja2ChatFormatter.IgnoreGenerationTags,
                jinja2.ext.loopcontrols,
            ],
        )
        template_env.filters["tojson"] = Jinja2ChatFormatter.tojson
        template = template_env.from_string(self._get_chat_template(llama))

        def raise_exception(message: str):
            raise ValueError(message)