Dataset languages
Bachstelze opened this issue · comments
There are many languages described in the paper.
Is this the dataset for all of them?
This repo contains the relabeled targets for English, German and Russian. For pre-training, we used a Common Crawl dataset with 101 languages.