Investigating the potential of Sparse Mixtures-of-Experts for multi-domain neural machine translation

Open in new window