wolof-audio-data
35,000 merged lines ALFFA, FLEURS, bus
Official site consultedAI: datasets
Last verified: 1 October 2026
Wolof audio dataset of 35,075 rows (6.47 GB) for speech recognition. It merges four sources: ALFFA, FLEURS, Urban Bus Wolof Speech and Kallaama. Each entry has 16 kHz audio, a transcription and a source identifier.
Services
- 28,807 training and 6,268 test examples
- WAV or MP3 audio with Wolof transcription
Access requirements
Free download on Hugging Face, Apache 2.0 license.
Sources
- wolof-audio-data (new tab), verified on 1 October 2026Official site consulted