This directory contains language-specific data files. Most importantly, you will find here:
- A list of unique characters for the target language (e.g. English) in data/alphabet.txt
- A scorer package (data/lm/kenlm.scorer) generated with data/lm/generate_package.py. The scorer package includes a binary n-gram language model generated with data/lm/generate_lm.py.
For more information on how to build these resources from scratch, see data/lm/README.md