Module: tfmot.clustering.keras

Module containing clustering code built on Keras abstractions.


clusterable_layer module: Clusterable layer API class for Keras models.

experimental module


class CentroidInitialization: Specifies how the cluster centroids should be initialized.

class ClusterableLayer: Abstract Base Class for making your own keras layer clusterable.

class ClusteringAlgorithm: Class to implement highly efficient vectorised look-ups.

class ClusteringSummaries: Helper class to create tensorboard summaries for the clustering progress.


cluster_scope(...): Provides a scope in which Clustered layers and models can be deserialized.

cluster_weights(...): Modifies a keras layer or model to be clustered during training.

strip_clustering(...): Strips clustering wrappers from the model.