diff --git a/README_CN.md b/README_CN.md
index 65a43ef..f2c4e01 100644
--- a/README_CN.md
+++ b/README_CN.md
@@ -82,9 +82,9 @@ from transformers import AutoModelForCausalLM, AutoTokenizer
 from transformers.generation import GenerationConfig
 
 tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen-7B-Chat", trust_remote_code=True)
-## 使用bf16精度
+## 打开bf16精度，A100、H100、RTX3060、RTX3070等显卡建议启用以节省显存
 # model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen-7B-Chat", device_map="auto", trust_remote_code=True, bf16=True).eval()
-## 使用fp16精度
+## 打开fp16精度，V100、P100、T4等显卡建议启用以节省显存
 # model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen-7B-Chat", device_map="auto", trust_remote_code=True, fp16=True).eval()
 # 默认使用fp32精度
 model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen-7B-Chat", device_map="auto", trust_remote_code=True).eval()
@@ -120,9 +120,9 @@ from transformers import AutoModelForCausalLM, AutoTokenizer
 from transformers.generation import GenerationConfig
 
 tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen-7B", trust_remote_code=True)
-## 使用bf16精度
+## 打开bf16精度，A100、H100、RTX3060、RTX3070等显卡建议启用以节省显存
 # model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen-7B", device_map="auto", trust_remote_code=True, bf16=True).eval()
-## 使用fp16精度
+## 打开fp16精度，V100、P100、T4等显卡建议启用以节省显存
 # model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen-7B", device_map="auto", trust_remote_code=True, fp16=True).eval()
 # 默认使用fp32精度
 model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen-7B", device_map="auto", trust_remote_code=True).eval()