Nesciun Lengaz Lascià Endò: Machine Translation for Fassa Ladin

A comparative analysis of the results obtained is reported, showing that jointly training for multilingual translation (Ladin-Italian and Ladin-English) significantly improves the performance, and knowledge-transfer is highly effective, highlighting the importance of targeted data collection and model adaptation in the context of low-resource/endangered languages for which little textual data is available.