Sparse communication for distributed gradient descent


We make distributed stochastic gradient descent faster by exchanging sparse updates instead of dense updates. Gradient updates are positively skewed as most updates are near zero, so we map the 99% smallest updates (by absolute value) to zero then exchange sparse matrices. This method can be combined with quantization to further improve the compression. We explore different configurations and apply them to neural machine translation and MNIST image classification tasks. Most configurations work on MNIST, whereas different configurations reduce convergence rate on the more complex translation task. Our experiments show that we can achieve up to 49% speed up on MNIST and 22% on NMT without damaging the final accuracy or BLEU.

Citation information

Aji, A & Heafield, K 2017, Sparse Communication for Distributed Gradient Descent. in Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing (EMNLP 2017). Association for Computational Linguistics (ACL), pp. 440-445, EMNLP 2017: Conference on Empirical Methods in Natural Language Processing, Copenhagen, Denmark, 7-11 September.