Stable Parallel Training of Wasserstein Conditional Generative Adversarial Neural Networks

TL;DR

This paper introduces a stable, scalable parallel training method for Wasserstein Conditional GANs that avoids inter-process communication, reduces mode collapse, and improves image quality on large datasets using extensive GPU resources.

Contribution

It presents a novel parallel training approach for W-CGANs that enhances stability and scalability without inter-process communication, outperforming previous methods.

Findings

Improved inception scores and Frechet distances over previous methods.

Achieved weak scaling on up to 2,000 GPUs.

Enhanced image quality in generated samples.

Abstract

We propose a stable, parallel approach to train Wasserstein Conditional Generative Adversarial Neural Networks (W-CGANs) under the constraint of a fixed computational budget. Differently from previous distributed GANs training techniques, our approach avoids inter-process communications, reduces the risk of mode collapse and enhances scalability by using multiple generators, each one of them concurrently trained on a single data label. The use of the Wasserstein metric also reduces the risk of cycling by stabilizing the training of each generator. We illustrate the approach on the CIFAR10, CIFAR100, and ImageNet1k datasets, three standard benchmark image datasets, maintaining the original resolution of the images for each dataset. Performance is assessed in terms of scalability and final accuracy within a limited fixed computational time and computational resources. To measure accuracy, we use the inception score, the Frechet inception distance, and image quality. An improvement in inception score and Frechet inception distance is shown in comparison to previous results obtained by performing the parallel approach on deep convolutional conditional generative adversarial neural networks (DC-CGANs) as well as an improvement of image quality of the new images created by the GANs approach. Weak scaling is attained on both datasets using up to 2,000 NVIDIA V100 GPUs on the OLCF supercomputer Summit.