Marian: Fast Neural Machine Translation in C++
Authors:
Marcin Junczys-Dowmunt,
Roman Grundkiewicz,
Tomasz Dwojak,
Hieu Hoang,
Kenneth Heafield,
Tom Neckermann,
Frank Seide,
Ulrich Germann,
Alham Fikri Aji,
Nikolay Bogoychev,
André F. T. Martins,
Alexandra Birch
Abstract:
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.
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.
△ Less
Submitted 4 April, 2018; v1 submitted 1 April, 2018;
originally announced April 2018.
