Developer reference

This section describes how to create NF architectures and NF components in Torchflows. NFs consist of two main components:

• a base distribution,

• a bijection.

In Torchflows, we further wrap these two with the torchflows.flows.Flow object or one of its subclasses to enable e.g., fitting NFs, computing the log probability density, and sampling.

At its core, each of these components is a PyTorch module which extends existing base classes:

• torch.distributions.Distribution and torch.nn.Module for base distributions,

• torchflows.bijections.base.Bijection for bijections,

• torchflows.flows.BaseFlow for flow wrappers.

Check the following pages for existing subclasses and to learn to create new subclasses for your modeling and research needs:

• Base distributions
• Bijections
• Flow wrappers