Timezone: »

Random Sharpness-Aware Minimization
Yong Liu · Siqi Mai · Minhao Cheng · Xiangning Chen · Cho-Jui Hsieh · Yang You

Wed Nov 30 09:00 AM -- 11:00 AM (PST) @ Hall J #701

Currently, Sharpness-Aware Minimization (SAM) is proposed to seek the parameters that lie in a flat region to improve the generalization when training neural networks. In particular, a minimax optimization objective is defined to find the maximum loss value centered on the weight, out of the purpose of simultaneously minimizing loss value and loss sharpness. For the sake of simplicity, SAM applies one-step gradient ascent to approximate the solution of the inner maximization. However, one-step gradient ascent may not be sufficient and multi-step gradient ascents will cause additional training costs. Based on this observation, we propose a novel random smoothing based SAM (R-SAM) algorithm. To be specific, R-SAM essentially smooths the loss landscape, based on which we are able to apply the one-step gradient ascent on the smoothed weights to improve the approximation of the inner maximization. Further, we evaluate our proposed R-SAM on CIFAR and ImageNet datasets. The experimental results illustrate that R-SAM can consistently improve the performance on ResNet and Vision Transformer (ViT) training.

Author Information

Yong Liu (National University of Singapore)
Siqi Mai (National University of Singapore)
Minhao Cheng (Hong Kong University of Science and Technology)
Xiangning Chen (UCLA, Google Brain)
Cho-Jui Hsieh (UCLA, Amazon)
Yang You (National University of Singapore)

More from the Same Authors