1x1 Conv1d with customized initialization.

Conv1d module with customized initialization.

PQMF module. This module is based on `Near-perfect-reconstruction pseudo-QMF banks`_. .. _`Near-perfect-reconstruction pseudo-QMF banks`:

Residual block module in WaveNet.

PyTorch compatible audio and mel dataset.

Customized collater for Pytorch DataLoader in training.

Creates an internal copy of a sequence and allows access to its items in a constrained random order. For a source sequence of n items and on

PyTorch compatible audio dataset.

CausalConv1d module with customized initialization.

Conv2d module with customized initialization.

Registers metadata about the dataset in a text file.

Log STFT magnitude loss module.

PyTorch compatible mel dataset.

Multi resolution STFT loss module.

STFT loss module.

Spectral convergence loss module.

Stretch2d module.

Tensorflow ReflectionPad1d module.

Customized trainer module for Multi-Singer training.

Upsampling network module.

CausalConvTranspose1d module with customized initialization.

Convolution + upsampling network module.

Customized collater for Pytorch DataLoader in training.

Generator1 module.

Generator2 module.

Loader class for a fests.scp file of hdf5 file. Examples: key1 /some/path/a.h5:feats key2 /some/path/b.h5:feats key3 /som

Loader class for a fests.scp file of npy file. Examples: key1 /some/path/a.npy key2 /some/path/b.npy key3 /some/path/c.np

Rectified Adam optimizer.

Residual block module in WaveNet.

Residual stack module introduced in MelGAN.

SingerConditional Discriminator module.

Tensorflow ConvTranspose1d module.

Tensorflow ResidualStack module.

Unconditional Discriminator module.