Translation/mttools

Using the mttools module

• Log into Taito or Abel
• Activate the NLPL software repository and load the module:

  module use -a /proj*/nlpl/software/modulefiles/
  module load nlpl-mttools

• Module-specific help is available by typing:

  module help nlpl-mttools

The following scripts are part of this module:

• moses-scripts
• Tokenization, casing, corpus cleaning and evaluation scripts from Moses
• Source: https://github.com/moses-smt/mosesdecoder (scripts directory)
• Installed revision: 413ba6b
• The subfolders generic, recaser, tokenizer, training are in PATH
• sacremoses
• Python port of Moses tokenizer and truecaser
• Source: https://github.com/alvations/sacremoses
• Installed version: 0.0.5
• subword-nmt
• Unsupervised Word Segmentation (a.k.a. Byte Pair Encoding) for Machine Translation and Text Generation
• Source: https://github.com/rsennrich/subword-nmt
• Installed version: 0.3.6
• The subword-nmt executable is in PATH
• sentencepiece
• Unsupervised text tokenizer for Neural Network-based text generation
• Source: https://github.com/google/sentencepiece
• Installed version: 0.1.6
• The spm_* executables are in PATH
• sacreBLEU
• Reference BLEU implementation that auto-downloads test sets and reports a version string to facilitate cross-lab comparisons
• Source: https://github.com/mjpost/sacreBLEU
• Installed version: 1.2.12
• The sacrebleu executable is in PATH
• scoring
• Script that makes it easy to score machine translation output using NIST's BLEU and NIST, TER, and METEOR, by Ken Heafield
• Source: https://kheafield.com/code/scoring.tar.gz
• Installed version: Sept 19, 2012
• The score.rb script is in PATH

Contact: Yves Scherrer, University of Helsinki, firstname.lastname@helsinki.fi