# Yorùbá Academic Wikipedia Q&A Dataset

This dataset contains ~5,000 instruction-style academic question–answer pairs
in **pure Yorùbá with tone marks**, derived from Yorùbá Wikipedia.

## Format
Each sample follows:
- instruction: academic question in Yorùbá
- input: empty
- output: academic answer in Yorùbá

## Intended Use
- Instruction tuning
- Low-resource language modeling
- Educational assistants

## Source
- Yorùbá Wikipedia (cleaned & deduplicated)