PC32

The dataset is a multilingual parallel corpus of 32 English-centric language pairs.

BibTex: