Hi! I'm Niels from the Hugging Face community science team. I've updated the dataset card to properly reflect the Vision-R1-cold dataset introduced in the Vision-R1 paper.
This PR:
- Adds the `image-text-to-text` task category metadata.
- Links the dataset to the official paper and GitHub repository.
- Replaces the generic framework README with dataset-specific information.
- Provides the LLaMA-Factory `dataset_info.json` configuration for usage as documented in the official repository.
