Marian: Fast Neural Machine Translation in C++

04/01/2018

∙

by   Marcin Junczys-Dowmunt, et al.

∙

0

∙

share

We present Marian, an efficient and self-contained Neural Machine Translation framework with an integrated automatic differentiation engine based on dynamic computation graphs. Marian is written entirely in C++. We describe the design of the encoder-decoder framework and demonstrate that a research-friendly toolkit can achieve high training and translation speed.