get_posterior

torchbayesian.bnn.utils.get_posterior(param: Tensor, posterior: str | Tuple[str, Dict[str, Any]]) → VariationalPosterior

Creates an instance of a ‘VariationalPosterior’ subclass to replace a given tensor. For use in ‘bnn.BayesianModule’.

Parameters

paramTensor

The tensor for which to initialize a variational posterior.

posteriorstr | Tuple[str, Dict[str, Any]]

The variational posterior to instantiate. Either the posterior’s name (str) or a tuple of the name and a dictionary of keyword arguments for instantiation.

Returns

posterior_instanceVariationalPosterior

The instantiated variational posterior.