6 datasets found

The dataset used for training and testing the machine translation systems.

This dataset is used to evaluate the performance of context-aware machine translation systems. It consists of English-Russian subtitles with varying levels of context.

The IWSLT17 dataset is a multilingual parallel corpus of 5 languages.

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

The WikiMatrix dataset is a multilingual dataset that contains parallel texts between English and other languages.

High-quality human translations from books, leveraging the induction bias that high-quality human translations are superior to machine-generated translations.