Update README.md

README.md

@@ -13,10 +13,12 @@ We write our prompts in the ChatML format.
 ### With vLLM (recommended for much faster inference)
 
 <details><summary>Install vLLM</summary>
+
+  [Reference](https://vllm.readthedocs.io/en/latest/getting_started/installation.html)
+  
 ```bash
   pip install vllm
 ```
-  [Reference](https://vllm.readthedocs.io/en/latest/getting_started/installation.html)
 </details>
 
 ```python