Fast Neural Machine Translation in C++
Перейти к файлу
Marcin Junczys-Dowmunt b69ad3ce39 add costType to alignment training 2018-08-27 17:55:43 +02:00
.vscode Scheduler::update() now reverts back to previous normalization (by #sentences) unless disp-label-counts; 2018-03-21 09:15:11 -07:00
cmake find static NCCL 2018-08-05 22:08:44 -07:00
contrib removed old (incorrect) VS files 2018-08-20 13:22:18 -07:00
doc Added more references 2016-10-06 10:25:53 -05:00
scripts Remove leftover references to Python bindings 2018-06-06 20:04:47 +02:00
src add costType to alignment training 2018-08-27 17:55:43 +02:00
vs fixed Visual Studio projects to actually build the EXE 2018-08-20 14:03:43 -07:00
.clang-format Update clang-format 2017-10-28 17:02:15 +00:00
.gitignore bug fix: History::Add() should obey the actual EOS symbol from the given vocabulary; 2018-06-21 13:34:19 -07:00
CHANGELOG.md Update CHANGELOG 2018-08-17 13:16:06 +01:00
CMakeLists.txt add build type Nonative 2018-08-17 17:39:40 -07:00
CONTRIBUTING.md Update CONTRIBUTING 2017-11-13 12:29:53 +00:00
Doxyfile.in Exclude 3rd party code from generated docs 2016-10-03 12:31:54 -05:00
LICENSE.md Update LICENSE.md 2017-02-27 01:16:42 +00:00
README.md Update README.md 2018-07-02 23:56:43 -07:00
VERSION set version to 1.6.0 2018-08-08 16:21:04 -07:00

README.md

Marian

Join the chat at https://gitter.im/marian-nmt Build Status CPU Build Status Tests Status Twitter

Marian is a C++ GPU-specific parallel automatic differentiation library with operator overloading. It is the training framework used in the Marian toolkit. This repository is the development repo of https://github.com/marian-nmt/marian, use it at your own risk.

https://github.com/marian-nmt/marian is updated with stable versions of this repository.

Named in honour of Marian Rejewski, a Polish mathematician and cryptologist.

Compilation

cd marian-dev
mkdir -p build
cd build
cmake .. -DCMAKE_BUILD_TYPE=Release
make -j

Website

More information on https://marian-nmt.github.io

Mailing lists

Contributions

See CONTRIBUTING.md

Acknowledgements

The development of Marian received funding from the European Union's Horizon 2020 Research and Innovation Programme under grant agreements 688139 (SUMMA; 2016-2019), 645487 (Modern MT; 2015-2017), 644333 (TraMOOC; 2015-2017), 644402 (HiML; 2015-2017), the Amazon Academic Research Awards program, the World Intellectual Property Organization, and is based upon work supported in part by the Office of the Director of National Intelligence (ODNI), Intelligence Advanced Research Projects Activity (IARPA), via contract #FA8650-17-C-9117.

This software contains source code provided by NVIDIA Corporation.