SGDR: Stochastic Gradient Descent with Warm Restarts
This work addresses the challenge of optimizing deep neural networks for researchers and practitioners, offering an incremental improvement over existing gradient-based methods.
The paper tackled the problem of improving the anytime performance of stochastic gradient descent when training deep neural networks by proposing a simple warm restart technique, achieving new state-of-the-art results of 3.14% and 16.21% error rates on CIFAR-10 and CIFAR-100 datasets, respectively.
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