xlm-r (cross-lingual language model – roberta)

This model builds upon Facebook’s RoBERTa model for improved performance on cross-lingual classification tasks. It can handle multiple languages at the same time.