UnifiedQA-v2: Stronger Generalization via Broader Cross-Format Training

TL;DR

UnifiedQA-v2 is an improved question-answering model that leverages significantly more training data to enhance its ability to generalize across various formats and domains.

Contribution

It introduces a larger, more comprehensive training process for UnifiedQA, resulting in stronger generalization capabilities across multiple question-answering formats.

Findings

Improved in-domain performance

Enhanced cross-domain generalization

Utilizes roughly three times more datasets

Abstract

We present UnifiedQA-v2, a QA model built with the same process as UnifiedQA, except that it utilizes more supervision -- roughly 3x the number of datasets used for UnifiedQA. This generally leads to better in-domain and cross-domain results.