Suggestion: Different LLM for different agents

[daniel-counto]

Suggest to allow user to set different LLM for different agents. For example, GPT4 for coder agent, and GPT3.5-16k for agent that summarizes long articles.

[victordibia]

This is a good pattern.

Each agent can take an llm_config parameter.

Something like the below:

Have you tried assigning llm_configs this way yet? A notebook example here.

import os 
 
gpt4_llm_config = {
    "seed": 42,  # seed for caching and reproducibility
    "config_list": [{
        "model": "gpt-4", 
        "api_key": os.environ["OPENAI_API_KEY"],
    }],
    "temperature": 0,  # temperature for sampling
    "use_cache": True,  # whether to use cache
}

gpt35_llm_config = {
    "seed": 42,  # seed for caching and reproducibility
    "config_list": [{
        "model": "gpt-3.5-turbo-16k",
        "api_key": os.environ["OPENAI_API_KEY"],
    }],
    "temperature": 0,  # temperature for sampling
    "use_cache": True,  # whether to use cache
}

# create an AssistantAgent named "assistant"
assistant = autogen.AssistantAgent(
    name="coding assistant",
    llm_config= gpt35_llm_config,
)
# create a UserProxyAgent instance named "user_proxy"
user_proxy = autogen.UserProxyAgent(
    name="user_proxy",
    human_input_mode="NEVER",
    llm_config= gpt4_llm_config,
    max_consecutive_auto_reply=10,
    is_termination_msg=lambda x: x.get("content", "").rstrip().endswith("TERMINATE"),
    code_execution_config={
        "work_dir": "coding",
        "use_docker": False,  # set to True or image name like "python:3" to use docker
    },
)

[daniel-counto]

@victordibia thanks