This PR adds relevant metadata to the model card, including:
- **pipeline_tag**: reinforcement-learning
- **library_name**: transformers
- **license**: apache-2.0
It also improves the README by providing direct links to the paper and the official GitHub repository, a summary of the Agent-RRM framework, and the proper citation format for research use.
