Function prepare_request_resources

llama_cpp/server/app.py:158–188

(
    body: CreateCompletionRequest | CreateChatCompletionRequest,
    llama_proxy: LlamaProxy,
    body_model: str | None,
    kwargs,
)

def prepare_request_resources(
    body: CreateCompletionRequest | CreateChatCompletionRequest,
    llama_proxy: LlamaProxy,
    body_model: str | None,
    kwargs,
) -> llama_cpp.Llama:
    if llama_proxy is None:
        raise HTTPException(
            status_code=status.HTTP_503_SERVICE_UNAVAILABLE,
            detail="Service is not available",
        )
    llama = llama_proxy(body_model)
    if body.logit_bias is not None:
        kwargs["logit_bias"] = (
            _logit_bias_tokens_to_input_ids(llama, body.logit_bias)
            if body.logit_bias_type == "tokens"
            else body.logit_bias
        )

    if body.grammar is not None:
        kwargs["grammar"] = llama_cpp.LlamaGrammar.from_string(body.grammar)

    if body.min_tokens > 0:
        _min_tokens_logits_processor = llama_cpp.LogitsProcessorList(
            [llama_cpp.MinTokensLogitsProcessor(body.min_tokens, llama.token_eos())]
        )
        if "logits_processor" not in kwargs:
            kwargs["logits_processor"] = _min_tokens_logits_processor
        else:
            kwargs["logits_processor"].extend(_min_tokens_logits_processor)
    return llama


async def get_event_publisher(