Improve model card: Add pipeline tag and update paper link

This PR improves the model card by:
1. Adding the `pipeline_tag: image-text-to-text` to the metadata, which makes the model easier to find at https://huggingface.co/models?pipeline_tag=image-text-to-text.
2. Updating the paper link from arXiv to the Hugging Face Papers page (`https://huggingface.co/papers/2507.19427`) and changing the label to "Paper" for better accuracy and integration with the Hub ecosystem.

Everything else looks great!

README.md

@@ -1,7 +1,9 @@
 ---
-license: apache-2.0
 library_name: transformers
+license: apache-2.0
+pipeline_tag: image-text-to-text
 ---
+
 <div align="center">
   <picture>
       <img src="https://github.com/stepfun-ai/Step3/blob/main/figures/stepfun-logo.png?raw=true" width="30%" alt="StepFun: Cost-Effective Multimodal Intelligence">
@@ -27,7 +29,7 @@ library_name: transformers
 </div>
 
 <div align="center">
-<b>📰&nbsp;&nbsp;<a href="https://stepfun.ai/research/step3">Step3 Model Blog</a></b> &nbsp;&nbsp;&nbsp; | &nbsp;&nbsp;&nbsp; <b>📄&nbsp;&nbsp;<a href="https://arxiv.org/abs/2507.19427">Step3 System Blog</a></b>
+<b>📰&nbsp;&nbsp;<a href="https://stepfun.ai/research/step3">Step3 Model Blog</a></b> &nbsp;&nbsp;&nbsp; | &nbsp;&nbsp;&nbsp; <b>📄&nbsp;&nbsp;<a href="https://huggingface.co/papers/2507.19427">Paper</a></b>
 </div>
 
 ## Introduction