README.md

metadata

language:
  - fo
  - en
pretty_name: FO-EN Parallel
configs:
  - config_name: default
    data_files:
      - data.jsonl
  - config_name: ShareGPT_translations
    data_files:
      - sharegpt_translation.jsonl
license: other
task_categories:
  - text-generation

A large-scale Faroese–English parallel dataset built from human translations, dictionary sources, and high-quality synthetic pairs.

Note that a fraction of this has been sourced from: https://huggingface.co/datasets/barbaroo/Sprotin_parallel and https://huggingface.co/datasets/barbaroo/fo_en_synthetic