tff.types.TensorType

View source on GitHub

An implementation of tff.Type representing types of tensors in TFF.

Inherits From: Type

View aliases

Main aliases

tff.TensorType

tff.types.TensorType(
    dtype: _Dtype, shape: Optional[object] = None
)

Used in the notebooks

Used in the tutorials
Implementing Custom Aggregations Sending Different Data To Particular Clients With tff.federated_select Client-efficient large-model federated learning via `federated_select` and sparse aggregation Composing Learning Algorithms

Args
dtype	An instance of tf.dtypes.DType or one of the Numpy numeric types.
shape	An optional instance of tf.TensorShape or an argument that can be passed to its constructor (such as a list or a tuple). None yields the default scalar shape.

Raises
TypeError	if arguments are of the wrong types.

Attributes
dtype
shape

Methods

check_abstract

View source

check_abstract() -> None

Check that this is a tff.AbstractType.

check_assignable_from

View source

check_assignable_from(
    source_type: 'Type'
) -> None

Raises if values of source_type cannot be cast to this type.

check_equivalent_to

View source

check_equivalent_to(
    other: 'Type'
) -> None

Raises if values of 'other' cannot be cast to and from this type.

check_federated

View source

check_federated() -> None

Check that this is a tff.FederatedType.

check_function

View source

check_function() -> None

Check that this is a tff.FunctionType.

check_identical_to

View source

check_identical_to(
    other: 'Type'
) -> None

Raises if other and Type are not exactly identical.

check_placement

View source

check_placement() -> None

Check that this is a tff.PlacementType.

check_sequence

View source

check_sequence() -> None

Check that this is a tff.SequenceType.

check_struct

View source

check_struct() -> None

Check that this is a tff.StructType.

check_struct_with_python

View source

check_struct_with_python() -> None

Check that this is a tff.StructWithPythonType.

check_tensor

View source

check_tensor() -> None

Check that this is a tff.TensorType.

children

View source

children()

Returns a generator yielding immediate child types.

compact_representation

View source

compact_representation() -> str

Returns the compact string representation of this type.

formatted_representation

View source

formatted_representation() -> str

Returns the formatted string representation of this type.

is_abstract

View source

is_abstract() -> bool

Returns whether or not this type is a tff.AbstractType.

is_assignable_from

View source

is_assignable_from(
    source_type: tff.types.Type
) -> bool

Returns whether values of source_type can be cast to this type.

is_equivalent_to

View source

is_equivalent_to(
    other: 'Type'
) -> bool

Returns whether values of other can be cast to and from this type.

is_federated

View source

is_federated() -> bool

Returns whether or not this type is a tff.FederatedType.

is_function

View source

is_function() -> bool

Returns whether or not this type is a tff.FunctionType.

is_identical_to

View source

is_identical_to(
    other: 'Type'
) -> bool

Returns whether or not self and other are exactly identical.

is_placement

View source

is_placement() -> bool

Returns whether or not this type is a tff.PlacementType.

is_sequence

View source

is_sequence() -> bool

Returns whether or not this type is a tff.SequenceType.

is_struct

View source

is_struct() -> bool

Returns whether or not this type is a tff.StructType.

is_struct_with_python

View source

is_struct_with_python() -> bool

Returns whether or not this type is a tff.StructWithPythonType.

is_tensor

View source

is_tensor() -> bool

Returns whether or not this type is a tff.TensorType.

__eq__

View source

__eq__(
    other
)

Determines whether two type definitions are identical.

Note that this notion of equality is stronger than equivalence. Two types with equivalent definitions may not be identical, e.g., if they represent templates with differently named type variables in their definitions.

Args
other	The other type to compare against.

Returns
True if type definitions are syntactically identical (as defined above), otherwise False.

Raises
NotImplementedError	If not implemented in the derived class.

__ne__

View source

__ne__(
    other
)

Return self!=value.