Skip to content

Dataset · wmt

WMT English-German parallel corpus

wmt/wmt_t2t

Classic corpus of millions of English-German sentence pairs for machine translation tasks.

What's inside

A text dataset containing between 1M and 10M records prepared for the Tensor2Tensor library using data from statmt.org. It contains parallel texts in English and German from various years and sources. It also allows assembling custom datasets for other WMT language pairs.

What you can do with it

  • Training and fine-tuning machine translation models (English-German pair)
  • Evaluating translation system quality on standard WMT data
  • Pre-training bilingual language models
Tasks
Translation
Languages
German, English
Size
1M–10M rows
Downloads this month
22.8K
Likes
1

How to load it

from datasets import load_dataset

ds = load_dataset("wmt/wmt_t2t")