This PR significantly enhances the model card by:
- Adding `pipeline_tag: text-generation` to correctly classify the model's functionality and improve discoverability.
- Adding `library_name: transformers` to enable the automated "How to use" widget, as the model demonstrates compatibility with the `transformers` and `peft` libraries in its usage examples.
- Including direct links to the paper ([Principled RL for Diffusion LLMs Emerges from a Sequence-Level Perspective](https://huggingface.co/papers/2512.03759)), the project page ([https://jingyangou.github.io/ESPO-Demo/](https://jingyangou.github.io/ESPO-Demo/)), and the GitHub repository ([https://github.com/ML-GSAI/ESPO](https://github.com/ML-GSAI/ESPO)).
- Adding a "Sample Usage" section with a Python code snippet directly from the project's GitHub README, demonstrating how to load and use the LoRA adapter.
- Incorporating the paper's BibTeX citation for proper academic attribution.
