subword-nmt/README.md
2017-04-07 15:13:26 +02:00

1.4 KiB
Raw Blame History

Subword Neural Machine Translation

This repository contains preprocessing scripts to segment text into subword units. The primary purpose is to facilitate the reproduction of our experiments on Neural Machine Translation with subword units (see below for reference).

USAGE INSTRUCTIONS

Check the individual files for usage instructions.

To apply byte pair encoding to word segmentation, invoke these commands:

./learn_bpe.py -s {num_operations} < {train_file} > {codes_file}
./apply_bpe.py -c {codes_file} < {test_file}

To segment rare words into character n-grams, do the following:

./get_vocab.py < {train_file} > {vocab_file}
./segment-char-ngrams.py --vocab {vocab_file} -n {order} --shortlist {size} < {test_file}

The original segmentation can be restored with a simple replacement:

sed -r 's/(@@ )|(@@ ?$)//g'

PUBLICATIONS

The segmentation methods are described in:

Rico Sennrich, Barry Haddow and Alexandra Birch (2016): Neural Machine Translation of Rare Words with Subword Units Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (ACL 2016). Berlin, Germany.

ACKNOWLEDGMENTS

This project has received funding from Samsung Electronics Polska sp. z o.o. - Samsung R&D Institute Poland, and from the European Unions Horizon 2020 research and innovation programme under grant agreement 645452 (QT21).