Fast Neural Machine Translation in C++
Go to file
Marcin Junczys-Dowmunt d871ae674d Merge pull request #1 from gitter-badger/gitter-badge
Add a Gitter chat badge to README.md
2016-09-19 14:04:43 +01:00
examples merge 2016-09-16 11:48:03 +02:00
marian model dir 2016-09-15 14:17:04 +02:00
scripts attempts at relu and dropout nodes 2016-09-19 12:06:04 +02:00
src graphviz - BinaryNodeOp is a binary op 2016-09-19 13:26:54 +01:00
.gitignore add functions loading MNIST dataset 2016-09-13 18:13:23 +02:00
CMakeLists.txt Don't make doc when make all 2016-09-16 19:21:15 +02:00
Doxyfile.in Use CMake to invoke Doxygen 2016-09-16 18:28:00 +02:00
LICENSE.md Update LICENSE.md 2016-09-16 18:57:11 +02:00
README.md Add Gitter badge 2016-09-19 13:02:30 +00:00

Marian

Join the chat at https://gitter.im/MarianNMT/Lobby A C++ gpu-specific parallel automatic differentiation library with operator overloading.

In honour of Marian Rejewski, a Polish mathematician and cryptologist.

Installation

Requirements:

  • g++ with c++11
  • CUDA and CuDNN
  • Boost (>= 1.56)

Exporting some paths for CuDNN may be required (put it, for example, in your .bashrc file):

export PATH=$PATH:$HOME/.local/bin:/usr/local/cuda/bin
export LIBRARY_PATH=$LIBRARY_PATH:/usr/local/cuda/lib64:/usr/local/cuda/lib64/stubs:/usr/local/cudnn-5/lib64
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda/lib64:/usr/local/cuda/lib64/stubs:/usr/local/cudnn-5/lib64
export CPATH=$CPATH:/usr/local/cudnn-5/include

Compilation with cmake > 3.5:

mkdir build
cd build
cmake ..
make -j

To compile API documentation using Doxygen, first cd to the build directory, and then:

make doc