Improve model card: add pipeline tag, library name, and metadata

This PR improves the model card for HieroSA. It adds relevant metadata, including:
- `pipeline_tag: image-text-to-text`: This correctly identifies the model's multimodal task.
- `library_name: transformers`: The repository contains the standard `config.json` and associated files compatible with the Hugging Face Transformers library.

It also refines the formatting of the model card and ensures clear linking to the research paper and the official GitHub repository.

README.md

@@ -1,10 +1,13 @@
 ---
-license: mit
-language:
-- en
 base_model:
 - Qwen/Qwen3-VL-4B-Instruct
+language:
+- en
+license: mit
+pipeline_tag: image-text-to-text
+library_name: transformers
 ---
+
 # HieroSA (Chinese)
 
 [Paper](https://arxiv.org/abs/2601.05508) | [GitHub](https://github.com/THUNLP-MT/HieroSA)
@@ -15,17 +18,17 @@ HieroSA supports both modern logographic scripts and ancient hieroglyphs 🌍, e
 
 ## More Details
 
-Please refer to our [GitHub Repository](https://github.com/THUNLP-MT/HieroSA) for more details about this model.
+Please refer to our [GitHub Repository](https://github.com/THUNLP-MT/HieroSA) for more details about this model, including environment setup and inference scripts.
 
 ## Citation
 
 If you find our work helpful for your research, please consider citing our work.
 
-```plain
+```bibtex
 @article{luo2026hierosa,
     title={Enabling Stroke-Level Structural Analysis of Hieroglyphic Scripts without Language-Specific Priors}, 
     author={Fuwen Luo and Zihao Wan and Ziyue Wang and Yaluo Liu and Pau Tong Lin Xu and Xuanjia Qiao and Xiaolong Wang and Peng Li and Yang Liu},
     journal={arXiv preprint arXiv:2601.05508},
     year={2026}
 }
-```
+```