Hi!
The paper describes that the bitext is mined from the monolingual data. Is the cleaned monolingual data available anywhere?
· Sign up or log in to comment