Module: tff.learning.reconstruction

Libraries for using Federated Reconstruction algorithms.

Classes

class BatchOutput: A structure that holds the output of a tff.learning.reconstruction.Model.

class ClientOutput: Structure for outputs returned from clients during training.

class Model: Represents a reconstruction model for use in Tensorflow Federated.

Functions

build_dataset_split_fn(...): Builds a DatasetSplitFn for Federated Reconstruction training/evaluation.

build_federated_evaluation(...): Builds a tff.Computation for evaluating a reconstruction Model.

build_training_process(...): Builds the IterativeProcess for optimization using FedRecon.

from_keras_model(...): Builds a tff.learning.reconstruction.Model from a tf.keras.Model.

get_global_variables(...): Gets global variables from a Model as ModelWeights.

get_local_variables(...): Gets local variables from a Model as ModelWeights.

simple_dataset_split_fn(...): An example of a DatasetSplitFn that returns the original client data.

Type Aliases

DatasetSplitFn: The central part of internal API.