tf.types.experimental.FunctionType

Represents the type of a TensorFlow callable.

tf.types.experimental.FunctionType(
    parameters=None, *, return_annotation, __validate_parameters__=True
)

FunctionType inherits from inspect.Signature which canonically represents the structure (and optionally type) information of input parameters and output of a Python function. Additionally, it integrates with the tf.function type system (tf.types.experimental.TraceType) to provide a holistic representation of the the I/O contract of the callable. It is used for:

  • Canonicalization and type-checking of Python input arguments
  • Type-based dispatch to concrete functions
  • Packing/unpacking structured python values to Tensors
  • Generation of structured placeholder values for tracing

parameters

return_annotation

Child Classes

class empty

Methods

bind

bind(
    *args, **kwargs
)

Get a BoundArguments object, that maps the passed args and kwargs to the function's signature. Raises TypeError if the passed arguments can not be bound.

bind_partial

bind_partial(
    *args, **kwargs
)

Get a BoundArguments object, that partially maps the passed args and kwargs to the function's signature. Raises TypeError if the passed arguments can not be bound.

from_builtin

@classmethod
from_builtin(
    func
)

Constructs Signature for the given builtin function.

Deprecated since Python 3.5, use Signature.from_callable().

from_callable

View source

@classmethod
from_callable(
    obj, *, follow_wrapped=True
)

Constructs Signature for the given callable object.

from_function

@classmethod
from_function(
    func
)

Constructs Signature for the given python function.

Deprecated since Python 3.5, use Signature.from_callable().

replace

replace(
    *, parameters=_void, return_annotation=_void
)

Creates a customized copy of the Signature. Pass 'parameters' and/or 'return_annotation' arguments to override them in the new copy.

__eq__

__eq__(
    other
)

Return self==value.