CORE-T: COherent REtrieval of Tables for Text-to-SQL
Loading...
Date
2026-01-16
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Description
We provide preprocessed text-to-SQL benchmarks for BIRD, SPIDER, MMQA, and BEAVER. For BIRD, SPIDER, and MMQA, we preprocess the datasets to follow our open-book setting by merging tables from multiple DBs, or question-specific schemas for MMQA, into a single retrieval corpus per benchmark. BEAVER is already released in an open-book format. For consistency with the other benchmarks, we convert BEAVER's original MySQL database files into SQLite database files. We provide the preprocessed retrieval data together with the corresponding SQL databases.
Citation
Endorsement
DFG Classification
Project(s)
Faculty
Collections
License
Except where otherwise noted, this license is described as CC BY-SA 4.0