Upload README.md

escape '#'

README.md

@@ -55,7 +55,7 @@ model = PeftModel.from_pretrained(model, adapter_id_2, token=HF_TOKEN)
 def generate_response(input):
   prompt = f"""### 指示
   {input}
-  ### 回答
+  \#\#\# 回答
   """
 
   tokenized_input = tokenizer.encode(prompt, add_special_tokens=False, return_tensors="pt").to(model.device)