Skip to main content

Agent Chat with Async Human Inputs

Open In Colab Open on GitHub

%pip install "autogen" chromadb sentence_transformers tiktoken pypdf nest-asyncio
import asyncio
from typing import Any, Callable, Dict, List, Optional, Tuple, Union

import nest_asyncio

from autogen import AssistantAgent
from autogen.agentchat.contrib.retrieve_user_proxy_agent import RetrieveUserProxyAgent
from autogen.agentchat.user_proxy_agent import UserProxyAgent
# Define an asynchronous function that simulates some asynchronous task (e.g., I/O operation)


async def my_asynchronous_function():
print("Start asynchronous function")
await asyncio.sleep(2) # Simulate some asynchronous task (e.g., I/O operation)
print("End asynchronous function")
return "input"


# Define a custom class `CustomisedUserProxyAgent` that extends `UserProxyAgent`


class CustomisedUserProxyAgent(UserProxyAgent):
# Asynchronous function to get human input
async def a_get_human_input(self, prompt: str) -> str:
# Call the asynchronous function to get user input asynchronously
user_input = await my_asynchronous_function()

return user_input

# Asynchronous function to receive a message

async def a_receive(
self,
message: Union[Dict, str],
sender,
request_reply: Optional[bool] = None,
silent: Optional[bool] = False,
):
# Call the superclass method to handle message reception asynchronously
await super().a_receive(message, sender, request_reply, silent)


class CustomisedAssistantAgent(AssistantAgent):
# Asynchronous function to get human input
async def a_get_human_input(self, prompt: str) -> str:
# Call the asynchronous function to get user input asynchronously
user_input = await my_asynchronous_function()

return user_input

# Asynchronous function to receive a message
async def a_receive(
self,
message: Union[Dict, str],
sender,
request_reply: Optional[bool] = None,
silent: Optional[bool] = False,
):
# Call the superclass method to handle message reception asynchronously
await super().a_receive(message, sender, request_reply, silent)
def create_llm_config(model, temperature, seed):
config_list = [
{
"model": "<model_name>",
"api_key": "<api_key>",
},
]

llm_config = {
"seed": int(seed),
"config_list": config_list,
"temperature": float(temperature),
}

return llm_config
nest_asyncio.apply()


async def main():
boss = CustomisedUserProxyAgent(
name="boss",
human_input_mode="ALWAYS",
max_consecutive_auto_reply=0,
code_execution_config=False,
)

assistant = CustomisedAssistantAgent(
name="assistant",
system_message="You will provide some agenda, and I will create questions for an interview meeting. Every time when you generate question then you have to ask user for feedback and if user provides the feedback then you have to incorporate that feedback and generate new set of questions and if user don't want to update then terminate the process and exit",
llm_config=create_llm_config("gpt-4", "0.4", "23"),
)

await boss.a_initiate_chat(
assistant,
message="Resume Review, Technical Skills Assessment, Project Discussion, Job Role Expectations, Closing Remarks.",
n_results=3,
)


await main() # noqa: F704
boss (to assistant):

Resume Review, Technical Skills Assessment, Project Discussion, Job Role Expectations, Closing Remarks.

--------------------------------------------------------------------------------
assistant (to boss):

1. Can you walk me through your resume, highlighting the most significant parts?
2. What technical skills do you possess that make you a strong candidate for this position?
3. Can you discuss a project you've worked on that you believe showcases your abilities?
4. What are your expectations for this job role?
5. Do you have any questions or concerns before we conclude this interview?

Can you please provide your feedback on these questions?

--------------------------------------------------------------------------------
Provide feedback to assistant. Press enter to skip and use auto-reply, or type 'exit' to end the conversation: exit