Effective Gradient Sample Size via Variation Estimation for Accelerating Sharpness aware Minimization

Open in new window