tf.SparseTensorSpec

View source on GitHub

Class SparseTensorSpec

Type specification for a tf.SparseTensor.

Aliases:

• Class tf.compat.v1.SparseTensorSpec
• Class tf.compat.v2.SparseTensorSpec

__init__

View source

__init__(
    shape=None,
    dtype=tf.dtypes.float32
)

Constructs a type specification for a tf.SparseTensor.

Args:

• shape: The dense shape of the SparseTensor, or None to allow any dense shape.
• dtype: tf.DType of values in the SparseTensor.

Properties

dtype

The tf.dtypes.DType specified by this type for the SparseTensor.

shape

The tf.TensorShape specified by this type for the SparseTensor.

value_type

Methods

tf.SparseTensorSpec.__eq__

View source

__eq__(other)

Return self==value.

tf.SparseTensorSpec.__ne__

View source

__ne__(other)

Return self!=value.

tf.SparseTensorSpec.from_value

View source

@classmethod
from_value(
    cls,
    value
)

tf.SparseTensorSpec.is_compatible_with

View source

is_compatible_with(spec_or_value)

Returns true if spec_or_value is compatible with this TypeSpec.

tf.SparseTensorSpec.most_specific_compatible_type

View source

most_specific_compatible_type(other)

Returns the most specific TypeSpec compatible with self and other.

Args:

• other: A TypeSpec.

Raises:

• ValueError: If there is no TypeSpec that is compatible with both self and other.