Update app.py

app.py

@@ -78,7 +78,7 @@ async def generate_text(request: GenerateRequest = None, prompt: str = None, max
         raw_response = tokenizer.decode(outputs[0], skip_special_tokens=True)
         
         clean_response = raw_response.replace(formatted_prompt, "").strip()
-        clean_response = re.sub(r'</?s>|\[/?INST\]|\[/?INSR\]|\{/?INSST\}', '', clean_response).strip()
+        clean_response = re.sub(r'</?s>|\[/?s\]|\[/?INST\]|\[/?INSR\]|\{/?INSST\}', '', clean_response).strip()
         
         return {"response": clean_response}
     except Exception as e: