[Bug]: Bug in litellm.utils for checking function calling

[kdziedzic68]

What happened?

        "functions" in non_default_params
        or "function_call" in non_default_params
        or "tools" in non_default_params
    ):
        if (
            custom_llm_provider == "ollama"
            and custom_llm_provider != "text-completion-openai"
            and custom_llm_provider != "azure"
            and custom_llm_provider != "vertex_ai"
            and custom_llm_provider != "anyscale"
            and custom_llm_provider != "together_ai"
            and custom_llm_provider != "groq"
            and custom_llm_provider != "nvidia_nim"
            and custom_llm_provider != "cerebras"
            and custom_llm_provider != "xai"
            and custom_llm_provider != "ai21_chat"
            and custom_llm_provider != "volcengine"
            and custom_llm_provider != "deepseek"
            and custom_llm_provider != "codestral"
            and custom_llm_provider != "mistral"
            and custom_llm_provider != "anthropic"
            and custom_llm_provider != "cohere_chat"
            and custom_llm_provider != "cohere"
            and custom_llm_provider != "bedrock"
            and custom_llm_provider != "ollama_chat"
            and custom_llm_provider != "openrouter"
            and custom_llm_provider not in litellm.openai_compatible_providers
        ):
            if custom_llm_provider == "ollama":
                # ollama actually supports json output
                optional_params["format"] = "json"
                litellm.add_function_to_prompt = (
                    True  # so that main.py adds the function call to the prompt
                )
                if "tools" in non_default_params:
                    optional_params["functions_unsupported_model"] = (
                        non_default_params.pop("tools")
                    )
                    non_default_params.pop(
                        "tool_choice", None
                    )  # causes ollama requests to hang
                elif "functions" in non_default_params:
                    optional_params["functions_unsupported_model"] = (
                        non_default_params.pop("functions")
                    )
            elif (
                litellm.add_function_to_prompt
            ):  # if user opts to add it to prompt instead
                optional_params["functions_unsupported_model"] = non_default_params.pop(
                    "tools", non_default_params.pop("functions", None)
                )
            else:
                raise UnsupportedParamsError(
                    status_code=500,
                    message=f"Function calling is not supported by {custom_llm_provider}.",
                )```

a second layer of conditions would be true only for `ollama` model - other providers would be validated positively

### Relevant log output

```shell

Are you a ML Ops Team?

Yes

What LiteLLM version are you on ?

I don't know - an issue is pointing to code on main so im assuming that newest

Twitter / LinkedIn details

No response

[krrishdholakia]

@kdziedzic68 I don't follow. What is the error and how do i reproduce it?

[kdziedzic68]

there is no error in a code execution, but in the logic of conditional statements.

please note that

if (
            custom_llm_provider == "ollama"
            and custom_llm_provider != "text-completion-openai"
            and custom_llm_provider != "azure"
            and custom_llm_provider != "vertex_ai"
            and custom_llm_provider != "anyscale"
            and custom_llm_provider != "together_ai"
            and custom_llm_provider != "groq"
            and custom_llm_provider != "nvidia_nim"
            and custom_llm_provider != "cerebras"
            and custom_llm_provider != "xai"
            and custom_llm_provider != "ai21_chat"
            and custom_llm_provider != "volcengine"
            and custom_llm_provider != "deepseek"
            and custom_llm_provider != "codestral"
            and custom_llm_provider != "mistral"
            and custom_llm_provider != "anthropic"
            and custom_llm_provider != "cohere_chat"
            and custom_llm_provider != "cohere"
            and custom_llm_provider != "bedrock"
            and custom_llm_provider != "ollama_chat"
            and custom_llm_provider != "openrouter"
            and custom_llm_provider not in litellm.openai_compatible_providers
        ):

is equivalent to

if custom_llm_provider == "ollama":
 ....