AutoGen

Agents

an agent in autogen, is an entity that can sed and receive messages to and from other agents in its environment.
An agent can be powered by models, code executors (such as an IPyhon kernel), human, or a combination of these and other pluggable and customizable components.

An example of such agents is the built-in ConversableAgent which supports llm, code executor, function and tool executor, a component for keeping human-in-the-loop


from autogen import ConversableAgent, config_list_from_json

config_list = config_list_from_json(env_or_file="OAI_CONFIG_LIST")

agent = ConversableAgent(
    "chatbot",
    llm_config={"config_list": config_list},
    code_execution_config=False, # turn off code executor default off
    function_map=None, # no registerd functions, default none
    human_input_mode="NEVER" # never ask for human input,

)

reply = agent.generate_reply(messages=[
    {
        "content": "tell me a joke",
        "role": "user"
    }
])

print(reply)

Roles and Conversation

assgin roles to agents and have them participate in conversations or chat with each other.
A conversation is a sequence of messages exchanged between agents.
for example, we can assign different roles to two agents by setting their system_message.


cathy = ConversableAgent(
    "cathy",
    llm_config={"config_list": config_list},
    system_message="Your name is Cathy and your are a part of a duo of comedians.",
    human_input_mode="NEVER"
)

joe = ConversableAgent(
    "joe",
    system_message="Your name is Joe and you are a part of a duo of comedians.",
    llm_config={"config_list": config_list},
    human_input_mode="NEVER"
)

result = joe.initiate_chat(cathy, message="Cathy, tell me a joke", max_turns=2)


joe (to cathy):

Cathy, tell me a joke

--------------------------------------------------------------------------------
cathy (to joe):

Sure, Joe! Why did the scarecrow win an award?  

Because he was outstanding in his field! 🌾😄

--------------------------------------------------------------------------------
joe (to cathy):

That's a classic! Alright, how about this: Why don't skeletons fight each other? Because they don't have the guts! 🦴🤣

--------------------------------------------------------------------------------
cathy (to joe):

Good one, Joe! Alright, here’s another: What do you call a fish wearing a bowtie? 

Sofishticated! 🎩🐟

总结

主要是初步了解是试用autogen，但这个工具和一年前我了解的metagpt很类似，就是角色环境工具使用，但autogen在这方面的设计应该更灵活一些，最近也是有一些想法想要通过mutiple-agents系统来实现看看效果，目前简单的使用上没有什么问题，接下来想看看里面的其他的模块，比如工具使用，代码执行部分的内容，能够和图数据进行结合的话更好了。