Neural Nets

In this module are stored the main Neural Networks Architectures.

• Base Neural Nets
• Benchmarks

The Base Architectures

BaseEncoder	This is a base class for Encoders neural networks.
BaseDecoder	This is a base class for Decoders neural networks.
BaseDiscriminator	This is a base class for Discriminator neural networks.
BaseMetric	This is a base class for Metrics neural networks (only applicable for Riemannian based VAE)

The Benchmark Architectures

In this model are gathered some predefined neural nets architectures that may be used on benchmark datasets such as MNIST, CIFAR or CELEBA

MNIST

A collection of Neural nets used to perform the benchmark on MNIST

ConvNets

Encoder_Conv_AE_MNIST	A Convolutional encoder suited for MNIST and Autoencoder-based models.
Encoder_Conv_VAE_MNIST	A Convolutional encoder suited for MNIST and Variational Autoencoder-based models.
Encoder_Conv_SVAE_MNIST	A Convolutional encoder suited for mnist and Hyperspherical autoencoder Variational Autoencoder.
Decoder_Conv_AE_MNIST	A Convolutional decoder suited for MNIST and Autoencoder-based models.
Discriminator_Conv_MNIST	A Convolutional discriminator suited for MNIST.

ResNets

Encoder_ResNet_AE_MNIST	A ResNet encoder suited for MNIST and Autoencoder-based models.
Encoder_ResNet_VAE_MNIST	A ResNet encoder suited for MNIST and Variational Autoencoder-based models.
Encoder_ResNet_SVAE_MNIST	A ResNet encoder suited for MNIST and Hyperspherical VAE models.
Encoder_ResNet_VQVAE_MNIST	A ResNet encoder suited for MNIST and Vector Quantized VAE models.
Decoder_ResNet_AE_MNIST	A ResNet decoder suited for MNIST and Autoencoder-based models.
Decoder_ResNet_VQVAE_MNIST	A ResNet decoder suited for MNIST and Vector Quantized VAE models.

CIFAR

A collection of Neural nets used to perform the benchmark on CIFAR

ConvNets

Encoder_Conv_AE_CIFAR	A Convolutional encoder Neural net suited for CIFAR and Autoencoder-based models.
Encoder_Conv_VAE_CIFAR	A Convolutional encoder Neural net suited for CIFAR and Variational Autoencoder-based models.
Encoder_Conv_SVAE_CIFAR	A Convolutional encoder Neural net suited for CIFAR and Hyperspherical Variational Autoencoder.
Decoder_Conv_AE_CIFAR	A Convolutional decoder Neural net suited for CIFAR and Autoencoder-based models.
Discriminator_Conv_CIFAR	A Convolutional discriminator Neural net suited for CIFAR.

ResNets

Encoder_ResNet_AE_CIFAR	A ResNet encoder suited for CIFAR and Autoencoder-based models.
Encoder_ResNet_VAE_CIFAR	A ResNet encoder suited for CIFAR and Variational Autoencoder-based models.
Encoder_ResNet_SVAE_CIFAR	A ResNet encoder suited for CIFAR and Hyperspherical VAE models.
Encoder_ResNet_VQVAE_CIFAR	A ResNet encoder suited for CIFAR and Vector Quantized VAE models.
Decoder_ResNet_AE_CIFAR	A ResNet decoder suited for CIFAR and Autoencoder-based models.
Decoder_ResNet_VQVAE_CIFAR	A ResNet decoder suited for CIFAR and Vector Quantized VAE models.

CELEBA-64

A collection of Neural nets used to perform the benchmark on CELEBA

ConvNets

Encoder_Conv_AE_CELEBA	A Convolutional encoder Neural net suited for CELEBA-64 and Autoencoder-based models.
Encoder_Conv_VAE_CELEBA	A Convolutional encoder Neural net suited for CELEBA-64 and Variational Autoencoder-based models.
Encoder_Conv_SVAE_CELEBA	A Convolutional encoder Neural net suited for CELEBA-64 and Hyperspherical autoencoder Variational Autoencoder.
Decoder_Conv_AE_CELEBA	A Convolutional decoder Neural net suited for CELEBA-64 and Autoencoder-based models.
Discriminator_Conv_CELEBA	A Convolutional discriminator Neural net suited for CELEBA.

ResNets

Encoder_ResNet_AE_CELEBA	A ResNet encoder suited for CELEBA and Autoencoder-based models.
Encoder_ResNet_VAE_CELEBA	A ResNet encoder suited for CELEBA and Variational Autoencoder-based models.
Encoder_ResNet_SVAE_CELEBA	A ResNet encoder suited for CELEBA and Hyperspherical VAE models.
Encoder_ResNet_VQVAE_CELEBA	A ResNet encoder suited for CELEBA and Vector Quantized VAE models.
Decoder_ResNet_AE_CELEBA	A ResNet decoder suited for CELEBA and Autoencoder-based models.
Decoder_ResNet_VQVAE_CELEBA	A ResNet decoder suited for CELEBA and Vector Quantized VAE models.

Note

In case you want to provide your own neural architecture, make sure you make them inherit from these classes. See tutorials.