Update main.py

main.py

@@ -20,5 +20,5 @@ app = FastAPI()
 # Endpoint for generating responses
 @app.post("/generate_response")
 async def generate_response(item: Validation):
-    prompt = f"<|start_header_id|>system<|end_header_id|> \n {item.system_prompt}<|eot_id|> \n <|start_header_id|>user<|end_header_id|>{item.user_prompt} <|eot_id|><|start_header_id|>assistant<|end_header_id|>"
+    prompt = f"<s>[INST] <<SYS>> \n {item.system_prompt}<</SYS>> \n {item.user_prompt} [/INST]"
     return llm.generate(prompt, do_sample=True)  # Adjusted to include the generation method with do_sample if needed