Skip to content

Dataset · Helsinki-NLP

Parallel corpus of European Parliament proceedings

Helsinki-NLP/europarl

A massive multilingual corpus of European Parliament speech translations for machine translation.

What's inside

A parallel text corpus ranging from 100 million to 1 billion lines, collected from the European Parliament website by Philipp Koehn from the University of Edinburgh. The data is available in European languages, including English, German, Bulgarian, Czech, Danish, and Greek.

What you can do with it

  • Training and evaluating machine translation models between European languages
  • Building cross-lingual word embeddings (CWLE)
Tasks
Translation
Languages
Bulgarian, Czech, Danish, German, Greek, English, Spanish, Estonian and 4 more
Size
100M–1B rows
Downloads this month
7K
Likes
41

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("Helsinki-NLP/europarl")