Update README.md

README.md

@@ -16,6 +16,6 @@ pinned: true
 Our mission is to promote knowledge of large language model training through real-world practices. By embracing openness and collaboration, we aim to advance Taiwan’s development in the field of generative AI.
 
 ### Open-Source Projects
-- **[Twinkle Eval](https://github.com/ai-twinkle/Eval)**: A fast and accurate AI evaluation tool that uses parallel and randomized testing methods to provide objective performance analysis and stability assessments. It supports various commonly used benchmark datasets.
+- 📊 **[Twinkle Eval](https://github.com/ai-twinkle/Eval)**: A fast and accurate AI evaluation tool that uses parallel and randomized testing methods to provide objective performance analysis and stability assessments. It supports various commonly used benchmark datasets.
 
 👋 Join our [Discord](https://discord.gg/Cx737yw4ed) to connect and collaborate with the community!