Module: tf.keras.utils

TensorFlow 1 version

Public API for tf.keras.utils namespace.


class CustomObjectScope: Exposes custom classes/functions to Keras deserialization internals.

class GeneratorEnqueuer: Builds a queue out of a data generator.

class HDF5Matrix: Representation of HDF5 dataset to be used instead of a Numpy array.

class OrderedEnqueuer: Builds a Enqueuer from a Sequence.

class Progbar: Displays a progress bar.

class Sequence: Base object for fitting to a sequence of data, such as a dataset.

class SequenceEnqueuer: Base class to enqueue inputs.

class custom_object_scope: Exposes custom classes/functions to Keras deserialization internals.


convert_all_kernels_in_model(...): Converts all convolution kernels in a model from Theano to TensorFlow. (deprecated)

deserialize_keras_object(...): Turns the serialized form of a Keras object back into an actual object.

get_custom_objects(...): Retrieves a live reference to the global dictionary of custom objects.

get_file(...): Downloads a file from a URL if it not already in the cache.

get_registered_name(...): Returns the name registered to an object within the Keras framework.

get_registered_object(...): Returns the class associated with name if it is registered with Keras.

get_source_inputs(...): Returns the list of input tensors necessary to compute tensor.

model_to_dot(...): Convert a Keras model to dot format.

multi_gpu_model(...): Replicates a model on different GPUs. (deprecated)

normalize(...): Normalizes a Numpy array.

pack_x_y_sample_weight(...): Packs user-provided data into a tuple.

plot_model(...): Converts a Keras model to dot format and save to a file.

register_keras_serializable(...): Registers an object with the Keras serialization framework.

serialize_keras_object(...): Serialize a Keras object into a JSON-compatible representation.

to_categorical(...): Converts a class vector (integers) to binary class matrix.

unpack_x_y_sample_weight(...): Unpacks user-provided data tuple.