R2D2-Net: Shrinking Bayesian Neural Networks via R2D2 Prior

Code repository for "R2D2-Net: Shrinking Bayesian Neural Networks via R2D2 Prior"

R2D2 Layers

The implementations of R2D2 layers together with other BNN layers can be found in the ./configs directory.

Architectures

The neural network architectures (e.g., LeNet) composed by the BNN layers can be found in the ./models directory. Users can customize own architectures also.

Get Started

Create a yaml config file in the ./configs directory (examples can be found in the same directory), and run the following codes to run an experiment

python main.py

Results will be saved in `./checkpoints'.