API Documentation¶
Core classes¶
Implementation of a variational autoencoder in TensorFlow
Dataset (*arrays, batch_size[, shuffle, to_dense]) |
Class for generating data mini-batches |
Model (n_inputs[, n_labels, optimizer, …]) |
Base abstract class for models |
UnsupervisedModel (n_inputs[, n_labels, …]) |
Base abstract class for unsupervised models |
VAE (n_inputs, n_latent[, n_encoder, …]) |
Variational Autoencoder |
Layers¶
TensorFlow layers
layers.dense (inputs, units[, activation, …]) |
Dense layer with optional weight normalization |
Operations¶
TensorFlow operations
ops.gaussian_log_likelihood (x[, mean, …]) |
Compute the log-likelihood for independent Gaussian variables |
ops.log_eluplusone (x[, name]) |
Compute log(elu(x) + 1) in a numerically stable manner |
ops.reduce_logmeanexp (x[, axis, keep_dims, name]) |
Compute log(sum(exp(x))) in a numerically stable manner |