Create a convolution filter variable with the specified name and shape, and initialize it using Xavier initialition.

Create a bias variable with the specified name and shape and initialize it to zero.

Perform convolution for a single convolutional processing step.

Construct the WaveNet network.

One-hot encodes the waveform amplitudes. This allows the definition of the network as a categorical distribution over a finite set of

Creates a single causal convolution layer. The layer can change the number of channels.

Creates a single causal dilated convolution layer. The layer contains a gated filter that connects to dense output and to a skip conn

Construct an efficient incremental generator.

This function creates all variables used by the network. This allows us to share them between multiple calls to the loss function and

Recursively finds all files matching the pattern.

Generator that yields text raw from the directory.

Creates a WaveNet network and returns the autoencoding loss. The variables are all scoped to the given name.

Computes the probability distribution of the next sample based on all samples in the input waveform. If you want to generate audio by

Computes the probability distribution of the next sample incrementally, based on a single sample and all previously passed samples.

Validate and arrange directory related arguments.

Initializes the WaveNet model. Args: batch_size: How many audio files are supplied per batch (recommended: 1).

Convert string to bool (in argparse context).