sfaira.models.embedding.ModelKerasVaeIAF

class sfaira.models.embedding.ModelKerasVaeIAF(in_dim, latent_dim=(128, 64, 2, 64, 128), n_iaf=2, dropout_rate=0.1, l2_coef: float = 0.0, l1_coef: float = 0.0, mc_samples=10, batchnorm=False, activation='tanh', init='glorot_uniform', output_layer='nb')

Attributes

version

Methods

predict_embedding(x[, variational, return_z0])
predict_reconstructed(x, **kwargs)