Skip to content

Dataset · AmazonScience

Mintaka Multilingual QA Dataset

AmazonScience/mintaka

Comprehensive dataset of 180,000 question-answer pairs in 9 languages for end-to-end QA systems.

What's inside

The dataset contains 20,000 English-language questions annotated with Wikidata entities, translated into Arabic, French, German, Hindi, Italian, Japanese, Portuguese, and Spanish. Includes 8 complex question types (including superlative, intersection, and multi-hop) collected from crowdworkers.

What you can do with it

  • Training and evaluating multilingual end-to-end QA models
  • Testing models on complex and multi-hop reasoning
Tasks
Question answering
Size
100K–1M rows
Downloads this month
2K
Likes
13

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("AmazonScience/mintaka")