Browse Datasets

Inkuba Instruct

Multi-task Instruction Data by Lelapa AI

instruction-tuning multi-task fine-tuning 62M Swahili, 6M Yoruba samples CC BY-NC 4.0
Access Dataset

Description

Developed by Lelapa AI, compiles downstream tasks for five African languages. Integrates MT, NER, POS, and Sentiment data for instruction fine-tuning.

Quick Start

Python 
from datasets import load_dataset

dataset = load_dataset("lelapa/Inkuba-instruct", "swahili_train", split="train")

Project Ideas

TinyLlama for Zulu

Fine-tune TinyLlama for Zulu instruction following

Advanced

Ethical Considerations

  • Non-commercial license