This PR improves the model card for **Innovator-VL-8B-Thinking** by:
1. Updating the `pipeline_tag` to `image-text-to-text` for better categorization.
2. Adding `library_name: transformers` to enable automated code snippets on the Hub.
3. Adding explicit links to the paper, project page, and GitHub repository.
4. Ensuring the model card provides a clear overview of the model's architecture and training stages as described in the technical report.
