tfr.keras.model.TypeSpecInputCreator

InputCreator with tensor type specs.

Inherits From: InputCreator

tfr.keras.model.TypeSpecInputCreator(
    type_spec: Dict[str, Union[tf.TensorSpec, tf.RaggedTensorSpec]],
    context_feature_names: Optional[List[str]] = None,
    example_feature_names: Optional[List[str]] = None
)

Example usage:

input_creator=TypeSpecInputCreator(
    {"example_feature_1": tf.TensorSpec(shape=[None, 1], dtype=tf.float32)},
    example_feature_names=["example_feature_1"])

Args
type_spec	A dict maps the context and example feature keys to the corresponding context and example type specs.
context_feature_names	A list of context feature keys.
example_feature_names	A list of example feature keys.

Methods

__call__

View source

__call__() -> Tuple[tfr.keras.model.TensorDict, tfr.keras.model.TensorDict]

See InputCreator.