PyMarian: Fast Neural Machine Translation and Evaluation in Python

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Gowda, Thamme, Grundkiewicz, Roman, Rippeth, Elijah, Post, Matt, Junczys-Dowmunt, Marcin
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914920498462720
author Gowda, Thamme
Grundkiewicz, Roman
Rippeth, Elijah
Post, Matt
Junczys-Dowmunt, Marcin
author_facet Gowda, Thamme
Grundkiewicz, Roman
Rippeth, Elijah
Post, Matt
Junczys-Dowmunt, Marcin
contents The deep learning language of choice these days is Python; measured by factors such as available libraries and technical support, it is hard to beat. At the same time, software written in lower-level programming languages like C++ retain advantages in speed. We describe a Python interface to Marian NMT, a C++-based training and inference toolkit for sequence-to-sequence models, focusing on machine translation. This interface enables models trained with Marian to be connected to the rich, wide range of tools available in Python. A highlight of the interface is the ability to compute state-of-the-art COMET metrics from Python but using Marian's inference engine, with a speedup factor of up to 7.8$\times$ the existing implementations. We also briefly spotlight a number of other integrations, including Jupyter notebooks, connection with prebuilt models, and a web app interface provided with the package. PyMarian is available in PyPI via $\texttt{pip install pymarian}$.
format Preprint
id arxiv_https___arxiv_org_abs_2408_11853
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle PyMarian: Fast Neural Machine Translation and Evaluation in Python
Gowda, Thamme
Grundkiewicz, Roman
Rippeth, Elijah
Post, Matt
Junczys-Dowmunt, Marcin
Computation and Language
The deep learning language of choice these days is Python; measured by factors such as available libraries and technical support, it is hard to beat. At the same time, software written in lower-level programming languages like C++ retain advantages in speed. We describe a Python interface to Marian NMT, a C++-based training and inference toolkit for sequence-to-sequence models, focusing on machine translation. This interface enables models trained with Marian to be connected to the rich, wide range of tools available in Python. A highlight of the interface is the ability to compute state-of-the-art COMET metrics from Python but using Marian's inference engine, with a speedup factor of up to 7.8$\times$ the existing implementations. We also briefly spotlight a number of other integrations, including Jupyter notebooks, connection with prebuilt models, and a web app interface provided with the package. PyMarian is available in PyPI via $\texttt{pip install pymarian}$.
title PyMarian: Fast Neural Machine Translation and Evaluation in Python
topic Computation and Language
url https://arxiv.org/abs/2408.11853