Update README.md

README.md

@@ -12,13 +12,15 @@ pipeline_tag: text-classification
 <hr>
 <div align="center" style="line-height: 1;">
   <a href="https://arxiv.org/abs/2507.01352" target="_blank">
-    <img alt="Paper" src="https://img.shields.io/badge/📖%20Paper-Skywork--Reward--V2-4D5EFF?style=flat-square&labelColor=202124"/>
+    <img alt="Paper" src="https://img.shields.io/badge/📖_Paper-Skywork--Reward--V2-4D5EFF?style=flat-square&labelColor=202124"/>
   </a>
   <a href="https://huggingface.co/collections/Skywork/skywork-reward-v2-685cc86ce5d9c9e4be500c84" target="_blank">
-    <img alt="Models" src="https://img.shields.io/badge/🤗_Hugging_Face-Skywork-4D5EFF?style=flat-square&labelColor=202124"/>
+    <img alt="Models" src="https://img.shields.io/badge/🤗_Hugging_Face-Model_Collection-4D5EFF?style=flat-square&labelColor=202124"/>
+  </a>
+  <a href="https://github.com/SkyworkAI/Skywork-Reward-V2" target="_blank">
+    <img alt="GitHub" src="https://img.shields.io/badge/🧑‍💻_GitHub-Skywork--Reward--V2-4D5EFF?style=flat-square&labelColor=202124"/>
   </a>
 </div>
-
 ## 🔥 Highlights
 
 **Skywork-Reward-V2** is a series of eight reward models designed for versatility across a wide range of tasks, trained on a mixture of 26 million carefully curated preference pairs. While the Skywork-Reward-V2 series remains based on the Bradley-Terry model, we push the boundaries of training data scale and quality to achieve superior performance. Compared to the first generation of Skywork-Reward, the Skywork-Reward-V2 series offers the following major improvements: