generative-ai-python

google.generativeai.types.CallableFunctionDeclaration

An extension of FunctionDeclaration that can be built from a python function, and is callable.

Inherits From: FunctionDeclaration

google.generativeai.types.CallableFunctionDeclaration(
    *,
    name: str,
    description: str,
    parameters: (dict[str, Any] | None) = None,
    function: Callable[..., Any]
)

Note: The python function must have type annotations.

Attributes

`description`
`name`
`parameters`

Methods

from_function

View source

@staticmethod
from_function(
    function: Callable[..., Any], descriptions: (dict[str, str] | None) = None
)

Builds a CallableFunctionDeclaration from a python function.

The function should have type annotations.

This method is able to generate the schema for arguments annotated with types:

AllowedTypes = float | int | str | list[AllowedTypes] | dict

This method does not yet build a schema for TypedDict, that would allow you to specify the dictionary contents. But you can build these manually.

from_proto

View source

@classmethod
from_proto(
    proto
) -> FunctionDeclaration

to_proto

View source

to_proto() -> protos.FunctionDeclaration

__call__

View source

__call__(
    fc: protos.FunctionCall
) -> protos.FunctionResponse

Call self as a function.

This site is open source. Improve this page.