This PR adds a comprehensive model card for the `3D-R1` model, which is presented in the paper [3D-R1: Enhancing Reasoning in 3D VLMs for Unified Scene Understanding](https://huggingface.co/papers/2507.23478).
The model card includes:
- A link to the paper.
- A link to the project page (`https://aigeeksgroup.github.io/3D-R1`).
- A link to the code repository (`https://github.com/AIGeeksGroup/3D-R1`).
- Relevant metadata: `pipeline_tag: image-text-to-text`, `library_name: transformers`, and `license: apache-2.0`.
This will help users discover and understand the model better on the Hugging Face Hub.
