This repository contains the code for the NeurIPS HITY workshop paper "SADT: Combining Sharpness-Aware Minimization with Self-Distillation for Improved Model Generalization". Provided code is the one-step solution to reproduce the results from the Paper.
