**License:** Apache-2.0

This dataset is derived from [flaviagiammarino/vqa-rad](https://huggingface.co/datasets/flaviagiammarino/vqa-rad).  
All credit goes to the original authors.

The dataset has been reformatted into `.tsv` files for compatibility with [VLMEvalKit](https://github.com/your-vlmevalkit-link).  
No changes were made to the original image or text content.

This version is intended for benchmarking vision-language models using VLMEvalKit.