View source on GitHub |
A no-op wrapper of datasets and signatures.
Inherits From: AbstractDatasetBuilder
tfr.keras.pipeline.NullDatasetBuilder(
train_dataset, valid_dataset, signatures=None
)
Example usage:
train_dataset = tf.data.Dataset(...)
valid_dataset = tf.data.Dataset(...)
dataset_builder = NullDatasetBuilder(train_dataset, valid_dataset)
Args | |
---|---|
train_dataset
|
A tf.data.Dataset for training.
|
valid_dataset
|
A tf.data.Dataset for validation.
|
signatures
|
A dict of signatures that formulate the model in functions that render the input data with given types. When None, no signatures assigned. |
Methods
build_signatures
build_signatures(
*arg, **kwargs
) -> Any
See AbstractDatasetBuilder
.
build_train_dataset
build_train_dataset(
*arg, **kwargs
) -> tf.data.Dataset
See AbstractDatasetBuilder
.
build_valid_dataset
build_valid_dataset(
*arg, **kwargs
) -> tf.data.Dataset
See AbstractDatasetBuilder
.