A dataset of over 3 million Tamazight(kabyle) text lines for NLP tasks.
International audience This study focuses on code-switching (CS) in French/Algerian A