SGDR: Stochastic Gradient Descent with Warm Restarts. Restart techniques are common in gradient-free optimization to deal with multimodal functions. Partial warm restarts are also gaining popularity in gradient-based optimization to improve the rate of convergence in accelerated gradient schemes to deal with ill-conditioned functions. In this paper, we propose a simple warm restart technique for stochastic gradient descent to improve its anytime performance when training deep neural networks. We empirically study its performance on the CIFAR-10 and CIFAR-100 datasets, where we demonstrate new state-of-the-art results at 3.14% and 16.21%, respectively. We also demonstrate its advantages on a dataset of EEG recordings and on a downsampled version of the ImageNet dataset. Our source code is available at https://github.com/loshchil/SGDR
Keywords for this software
References in zbMATH (referenced in 2 articles )
Showing results 1 to 2 of 2.
- Banert, Sebastian; Ringh, Axel; Adler, Jonas; Karlsson, Johan; Öktem, Ozan: Data-driven nonsmooth optimization (2020)
- Kaiyang Zhou, Tao Xiang: Torchreid: A Library for Deep Learning Person Re-Identification in Pytorch (2019) arXiv