File size: 460 Bytes
62634c2 6881ed7 62634c2 |
1 2 3 4 5 6 7 8 9 10 11 |
**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.
|