đź§ MetaWell Code-Switching Dataset
The MetaWell Code-Switching Dataset is the first large-scale African dataset built for training and evaluating code-switching between isiZulu and English.
It contains paired audio + text examples representing natural language use in South African daily life, media, and conversation. The dataset enables development of speech-to-text, translation, and language understanding models for African multilingual environments.
📊 Dataset Structure
Language Samples