chat = llm.predict({"inputs":prompt, "parameters":payload})
 
print(chat[0]["generated_text"])