G4KMU/

t2-ragbench

Project Page | Paper | Code

We deleted VQAonBD from the dataset due to low quality of the question reformulations. If you still want to use it you will find the data in the previous commit history.

	Dataset Description




	
	Dataset Summary

T2-RAGBench is a benchmark dataset designed to evaluate Retrieval-Augmented Generation (RAG) on financial documents containing both text and tables. It consists of 23,088… See the full description on the dataset page: https://huggingface.co/datasets/G4KMU/t2-ragbench.