文章预览
Messages：消息 上面调用LLM的方式，都是通过向 invoke 函数传入一个 string 参数，实现 text - in , text - out (文本输入，文本输出)，这是比较传统、老式的completion接口；现在LLM主推的API接口，都是所谓的Chat Completion Models方式，实现 messages in , message out （消息输入，消息输出）。直接看例子🌰： from langchain_openai import ChatOpenAI from langchain_core . messages import SystemMessage , HumanMessage load_dotenv () # LLM style: text in, text out model = ChatOpenAI ( model = "gpt-4o" ) result = model . invoke ( 'What is 81 divided by 9?' ) print ( "\n== LLM style: text in, text out ==\n" ) print ( result ) # SystemMessage: # Message for priming AI behavior, usually passed in as the first of a sequenc of input messages. # HumanMessagse: # Message from a human to the AI model. messages = [ SystemMessage ( content = "Solve the following math problems" ), HumanMessage ( c
………………………………
