poli.objective_repository.gfp_cbas.make_vae.BaseVAE

class poli.objective_repository.gfp_cbas.make_vae.BaseVAE(input_shape, latent_dim, *args, **kwargs)

Base class for Variational Autoencoders implemented in Keras

The class is designed to connect user-specified encoder and decoder models via a Model representing the latent space

__init__(input_shape, latent_dim, *args, **kwargs)

Methods

__init__(input_shape, latent_dim, *args, ...)
build_decoder(*args, **kwargs)	Build the decoder network as a keras Model
build_encoder(*args, **kwargs)	Build the encoder network as a keras Model
build_vae([epsilon_std, kl_scale])	Build the VAE
compile(*args, **kwargs)
fit(*args, **kwargs)
load_all_weights(prefix)
plot_model(*args, **kwargs)
save_all_weights(prefix)