This example compares the use of Pascal and Pytorch to write a simple Bayesian Neural Network that is trained using variational inference.
The PyTorch code was obtained from saxena-mayur/Weight-Uncertainty-in-Neural-Networks.
To run this example, run the following from the root directory of this project:
make eg-bnn-simple
Then look inside examples/bnn-simple/plots/ for the plots generated by Pascal and PyTorch. They should look exactly the same.