Marian: Cost-effective High-Quality Neural Machine Translation in C++
Microsoft Translator, 1 Microsoft Way, Redmond, WA 98121, USA
arXiv:1805.12096 [cs.CL], (30 May 2018)
@article{junczys-dowmunt2018marian,
title={Marian: Cost-effective High-Quality Neural Machine Translation in C++},
author={Junczys-Dowmunt, Marcin and Heafield, Kenneth and Hoang, Hieu and Grundkiewicz, Roman and Aue, Anthony},
year={2018},
month={may},
archivePrefix={"arXiv"},
primaryClass={cs.CL}
}
This paper describes the submissions of the "Marian" team to the WNMT 2018 shared task. We investigate combinations of teacher-student training, low-precision matrix products, auto-tuning and other methods to optimize the Transformer model on GPU and CPU. By further integrating these methods with the new averaging attention networks, a recently introduced faster Transformer variant, we create a number of high-quality, high-performance models on the GPU and CPU, dominating the Pareto frontier for this shared task.
June 2, 2018 by hgpu
