Parallel corpus

This dataset contains 70,861 English-Bangla sentence pairs and more than 0.8 million tokens in each side.