License: Apache-2.0
This dataset is derived from flaviagiammarino/vqa-rad.
All credit goes to the original authors.
The dataset has been reformatted into .tsv files for compatibility with VLMEvalKit.
No changes were made to the original image or text content.
This version is intended for benchmarking vision-language models using VLMEvalKit.