User Guide

For a quick start, take a look at the Examples.

The user guide contains more detailed information about using ProbFlow, including:

• A brief description of Bayesian modeling,

• Using Distributions, Parameters, and Modules to create Bayesian Models,

• How to fit those models to data,

• How to make predictions with those models,

• How to evaluate the performance of a model,

• How to inspect a model’s structure and the values of its parameters,

• How to use ProbFlow’s Ready-made Models,

• How to perform actions mid-training with Callbacks,

• How to load data on-the-fly with Data Generators,

• How to save and load models,

• How to choose your backend and default datatype, and

• Mathematical Details about how ProbFlow fits Bayesian models.