PhoMT: A High-Quality and Large-Scale Benchmark Dataset for Vietnamese-English Machine Translation

TL;DR

This paper introduces a large-scale Vietnamese-English dataset with 3.02 million sentence pairs and evaluates various translation models, demonstrating that fine-tuning mBART yields the best performance for this language pair.

Contribution

It provides the first large-scale Vietnamese-English dataset and benchmarks multiple models, highlighting the effectiveness of fine-tuning pre-trained auto-encoders.

Findings

Fine-tuning mBART achieves the best translation quality.

The dataset surpasses previous benchmarks in size and quality.

Both automatic and human evaluations favor mBART fine-tuning.

Abstract

We introduce a high-quality and large-scale Vietnamese-English parallel dataset of 3.02M sentence pairs, which is 2.9M pairs larger than the benchmark Vietnamese-English machine translation corpus IWSLT15. We conduct experiments comparing strong neural baselines and well-known automatic translation engines on our dataset and find that in both automatic and human evaluations: the best performance is obtained by fine-tuning the pre-trained sequence-to-sequence denoising auto-encoder mBART. To our best knowledge, this is the first large-scale Vietnamese-English machine translation study. We hope our publicly available dataset and study can serve as a starting point for future research and applications on Vietnamese-English machine translation.