SYSTRAN's Pure Neural Machine Translation Systems

TL;DR

This paper details SYSTRAN's development of production-ready neural machine translation systems across multiple languages, emphasizing practical choices, evaluation, and collaborative efforts to advance industry adoption.

Contribution

It presents a comprehensive approach to building and deploying large-scale NMT systems, including architecture, data handling, tuning, and evaluation, for real-world production use.

Findings

Successful deployment of NMT systems for 12 languages and 32 language pairs.

Evaluation methodology and initial performance insights shared.

Framework and practices aimed at industry adoption and further research.

Abstract

Since the first online demonstration of Neural Machine Translation (NMT) by LISA, NMT development has recently moved from laboratory to production systems as demonstrated by several entities announcing roll-out of NMT engines to replace their existing technologies. NMT systems have a large number of training configurations and the training process of such systems is usually very long, often a few weeks, so role of experimentation is critical and important to share. In this work, we present our approach to production-ready systems simultaneously with release of online demonstrators covering a large variety of languages (12 languages, for 32 language pairs). We explore different practical choices: an efficient and evolutive open-source framework; data preparation; network architecture; additional implemented features; tuning for production; etc. We discuss about evaluation methodology, present our first findings and we finally outline further work. Our ultimate goal is to share our expertise to build competitive production systems for "generic" translation. We aim at contributing to set up a collaborative framework to speed-up adoption of the technology, foster further research efforts and enable the delivery and adoption to/by industry of use-case specific engines integrated in real production workflows. Mastering of the technology would allow us to build translation engines suited for particular needs, outperforming current simplest/uniform systems.