The English and Faroese word or sentence pairs have been extracted from the Skype in Your Language project on Github. We then prepared them as a clean, aligned parallel dataset in json and csv. This resource provides bilingual data that can support localization and other development of language technology tools for Faroese.
There were two Faroese versions of the translations. Hence, the dataset contains two Faroese strings, fo1 and fo2. They are the same most of the time, but for some of the strings they differ.
The data is structured as follows:
Key, English, Faroese 1, Faroese 2, Similarity score, Similarity_label, Translation_coverage
Faroese strings: 3.452 (fo1), 3.196 (fo2)
Both fo1 and fo2: 3.178
Exact matches (fo1=fo2): 2.880
Release: 19.11.2025
Contact: mtd@setur.fo
