Skip to content

Latest commit

 

History

History
10 lines (6 loc) · 591 Bytes

File metadata and controls

10 lines (6 loc) · 591 Bytes

Bayesian-RRS-Models

This Repository contains some of the Bayesian Non Parametric Models Based on variational inference using Russian Roulette Sampling Method of approximating an infinite sum.

The papers reffered are mentioned here :

Variational Russian Roulette for Deep Bayesian Nonparametrics

Note : optimizer.py is taken from pytorch default library and is used for implementing dynamic version Adam optimizer for stochastic gumbel-softmax layers

APIC Net is a Bayesian Nueral Network modelling currently being developed using RRS VI.