Datasets:
metadata
task_categories:
- automatic-speech-recognition
- text-to-speech
language:
- lg
Luganda Bible Audio 100Hrs
- Audio split by chapters of Old & New Testament & Transcripts
- Format: 64kps, MP3
- Requires further splitting of the audio into smaller chunks/splits (https://github.com/facebookresearch/fairseq/tree/main/examples/mms/data_prep)
- Audio sourced from https://www.faithcomesbyhearing.com/