update README: reorder FlagOS and model download steps

README.md

@@ -64,17 +64,17 @@ We provide dequantized model weights in bfloat16 to run DeepSeek-R1 on NVIDIA GP
 
 ## 📌 Getting Started
 
-### Download open-source weights
+### Download the FlagOS image
 
 ```bash
-pip install modelscope
-modelscope download --model deepseek-ai/DeepSeek-R1 --local_dir /nfs/DeepSeek-R1
+docker pull flagrelease-registry.cn-beijing.cr.aliyuncs.com/flagrelease/flagrelease:deepseek-flagos-nvidia
 ```
 
-### Download the FlagOS image
+### Download open-source weights
 
 ```bash
-docker pull flagrelease-registry.cn-beijing.cr.aliyuncs.com/flagrelease/flagrelease:deepseek-flagos-nvidia
+pip install modelscope
+modelscope download --model deepseek-ai/DeepSeek-R1 --local_dir /nfs/DeepSeek-R1
 ```
 
 ### Start the inference service