MLQA

Parallel extractive question answering in seven languages, testing whether a model can recover an answer span from Wikipedia even when the question is in a different language.

Also known as: MultiLingual Question Answering, MLQA (MultiLingual Question Answering)

unassessed

This page is a discovery lead. Nobody has yet assessed it against the catalogue contract, so it carries no disposition. Absence of evidence here is not evidence of staleness.
Categoryknowledge
Subcategorymultilingual extractive question answering
Page statusactive
Metrictoken F1 and exact match, using the official MLQA evaluation script
Directionhigher_is_better
Unit%
Dataset size12738
Dataset licenceCC-BY-SA-3.0
PublisherFacebook AI Research

What it measures

MLQA (MultiLingual Question Answering) is an evaluation-only extractive QA set. The model receives a Wikipedia paragraph and a question, and must return the answer as a span in the paragraph's language. Seven languages are covered: English, Arabic, German, Spanish, Hindi, Vietnamese and Simplified Chinese. Instances are multi-way parallel, on average across four languages, so the same question can be asked against a context in another language. That is the generalised transfer setup. It is a reading-span task in SQuAD format, not a translation benchmark, and not the medical acronyms MEDIQA or MedQA.

Task format

Extractive span selection in SQuAD JSON. lm-evaluation-harness instead generates the answer string from a "Context / Question / Answer" prompt and scores it with the official MLQA normalisation (language-specific article stripping and Chinese mixed segmentation).

Models reporting this benchmark

No model card in ModelSpec reports this benchmark yet.

Data

This page as JSON · Edit on GitHub