Datasets:

---

honglyhly

/

DeepEyesV2_SFT

like 3

This PR adds a comprehensive dataset card for the DeepEyesV2 dataset.

It includes:

• The full abstract for DeepEyesV2: Toward Agentic Multimodal Model (https://huggingface.co/papers/2511.05271).
• Links to the associated paper, project page (https://visual-agent.github.io/), and code repository (https://github.com/Visual-Agent/DeepEyes).
• Metadata for task_categories (image-text-to-text), license (apache-2.0), and relevant tags for improved discoverability.
• A general usage description and the license information.