datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
Uganda-Multilingual-QA
BUAIR Uganda Multilingual Q&A
Parallel question–answer dataset for Ugandan languages, curated by the BUAIR Voice initiative at Busitema University.
Dataset version: v2 (updated 2026-08-31)
Each language config contains the same 4,256 agriculture / rural-livelihood Q&A pairs, with English as the shared source and translations into Japadhola, Ateso, Runyankore, and Luganda.
Changelog (v2)
Replaced v1 data (5,000 pairs from Multiligual-QA.xlsx) with cleaned data… See the full description on the dataset page: https://huggingface.co/datasets/BUAIR/Uganda-Multilingual-QA.buaiir_voice_jap
BUAIIR Japadhola Voice (BUAIR/buaiir_voice_jap)
Structured student read-speech in Japadhola (Adhola, ISO 639-3: adh) from Busitema University
Phase-2 batches (v, e, v2, e2). Maintained by BUAIIR.
Separate from Papoli community speech at BUAIR/popolivoice.
from datasets import load_dataset, Audio
ds = load_dataset("BUAIR/buaiir_voice_jap", split="train")
ds = ds.cast_column("audio", Audio(sampling_rate=16_000))
License: CC BY 4.0Updated: 2026-08-13
popolivoice
PopoliVoice (BUAIR/popolivoice)
Community-contributed Japadhola speech from the Papoli farming community (Tororo District, Uganda).
Maintained by BUAIIR — Busitema University AI & Innovation Research Lab.
This is the BUAIR org release. A copy is also available at Bateesa/popolivoice.
from datasets import load_dataset, Audio
ds = load_dataset("BUAIR/popolivoice", split="train")
ds = ds.cast_column("audio", Audio(sampling_rate=16_000))
License: CC BY 4.0Updated: 2026-08-13