Module: tfp.trainable_distributions

Trainable distributions.

"Trainable distributions" are instances of tfp.distributions which are parameterized by a transformation of a single input Tensor. The transformations are presumed to use TensorFlow variables and typically need to be fit, e.g., using tf.train optimizers or tfp.optimizers.


bernoulli(...): Constructs a trainable tfd.Bernoulli distribution.

multivariate_normal_tril(...): Constructs a trainable tfd.MultivariateNormalTriL distribution.

normal(...): Constructs a trainable tfd.Normal distribution.

poisson(...): Constructs a trainable tfd.Poisson distribution.

softplus_and_shift(...): Converts (batch of) scalars to (batch of) positive valued scalars.

tril_with_diag_softplus_and_shift(...): Converts (batch of) vectors to (batch of) lower-triangular scale matrices.

Other Members