This repo contains code that implements the following:
- An implementation of Adaptive Bayesian Optimization (ABO) using Sheffield Machine Learning's GPy and GPyOpt frameworks.
- A modified
torch.optim
the package that includes ABO implemented for the adaptive tuning of the learning rate (per epoch of training). - A set of experiments to test the efficacy of the adaptive learning rate tunning procedure.