tf.data.DatasetSpec

View source on GitHub

Type specification for tf.data.Dataset.

tf.data.DatasetSpec(
    element_spec, dataset_shape=()
)

See tf.TypeSpec for more information about TensorFlow type specifications.

>>> dataset = tf.data.Dataset.range(3)
>>> tf.data.DatasetSpec.from_value(dataset)
DatasetSpec(TensorSpec(shape=(), dtype=tf.int64, name=None), TensorShape([]))

Attributes:

• value_type: The Python type for values that are compatible with this TypeSpec.

Methods

__eq__

View source

__eq__(
    other
)

Return self==value.

__ne__

View source

__ne__(
    other
)

Return self!=value.

from_value

View source

@staticmethod
from_value(
    value
)

Creates a DatasetSpec for the given tf.data.Dataset value.

is_compatible_with

View source

is_compatible_with(
    spec_or_value
)

Returns true if spec_or_value is compatible with this TypeSpec.

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.