The amuNMT is a fast NMt decoder written in C++/CUDA. This image allows to download Rico Sennrich's models from WMT 2016 and run the decoder.
docker pull tomek/amunmt docker run -p 8756:8080 en-de
If you want to load for instance
en-ru model, replace
The input format is defined by swagger. You can find the swagger format file at