<ul><li>Researchers propose Asymptotic Unbiased Sampling with respect to iterations to accelerate Sharpness-Aware Minimization (SAM).</li><li>AUSAM probabilistically samples a subset of data points based on the Gradient Norm of each Sample (GNS).</li><li>The method consistently accelerates SAM across various tasks and networks, achieving a speedup of over 70%.</li><li>AUSAM is better suited for SAM and excels in maintaining performance compared to recent dynamic data pruning methods.</li></ul>

Asymptotic Unbiased Sample Sampling to Speed Up Sharpness-Aware Minimization

Discover more