Deep Architectures for Neural Machine Translation

TL;DR

This paper systematically compares various deep neural architectures for machine translation, introduces novel variants including BiDeep RNNs, and demonstrates improved translation quality and speed on English-German translation tasks.

Contribution

It provides a comprehensive evaluation of existing deep architectures, proposes the BiDeep RNN as a new approach, and shows its effectiveness in neural machine translation.

Findings

BiDeep RNN achieves 1.5 BLEU improvement over shallow baselines.

Several architectures improve translation quality and speed.

Deep models outperform shallower counterparts in translation tasks.

Abstract

It has been shown that increasing model depth improves the quality of neural machine translation. However, different architectural variants to increase model depth have been proposed, and so far, there has been no thorough comparative study. In this work, we describe and evaluate several existing approaches to introduce depth in neural machine translation. Additionally, we explore novel architectural variants, including deep transition RNNs, and we vary how attention is used in the deep decoder. We introduce a novel "BiDeep" RNN architecture that combines deep transition RNNs and stacked RNNs. Our evaluation is carried out on the English to German WMT news translation dataset, using a single-GPU machine for both training and inference. We find that several of our proposed architectures improve upon existing approaches in terms of speed and translation quality. We obtain best improvements with a BiDeep RNN of combined depth 8, obtaining an average improvement of 1.5 BLEU over a strong shallow baseline. We release our code for ease of adoption.