This dataset is a filtered sample of the Mozilla Common Voice 17.0 corpus, focusing on Swahili (sw)
A collection of read speech recordings in Swahili (Kiswahili).