Dataset · cardiffnlp
Tabular QA benchmark
cardiffnlp/databench
80 tabular datasets and questions in English and Spanish for LLM evaluation.
What's inside
An original corpus of 80 datasets prepared for the LREC-COLING 2024 paper and SemEval 2025 Task 8. The repository contains text data and table questions in English and Spanish, including test splits. The total volume ranges from 1 to 10 thousand records.
What you can do with it
- Evaluating the ability of language models to answer table-based questions
- Training or fine-tuning (SFT) models for structured data analysis
- Preparing and testing solutions for SemEval competition-level tasks
- Tasks
- Table question answering, Question answering
- Languages
- English, Spanish
- Size
- 1K–10K rows
- Downloads this month
- 24.8K
- Likes
- 16
Models trained on it
How to load it
from datasets import load_dataset
ds = load_dataset("cardiffnlp/databench")