mirror of https://github.com/microsoft/autogen.git
77 lines
2.7 KiB
Python
Executable File
77 lines
2.7 KiB
Python
Executable File
#!/usr/bin/env python3 -m pytest
|
|
|
|
import asyncio
|
|
import os
|
|
import sys
|
|
from unittest.mock import AsyncMock
|
|
|
|
import pytest
|
|
from test_assistant_agent import KEY_LOC, OAI_CONFIG_LIST
|
|
|
|
import autogen
|
|
|
|
sys.path.append(os.path.join(os.path.dirname(__file__), ".."))
|
|
from conftest import reason, skip_openai # noqa: E402
|
|
|
|
|
|
@pytest.mark.skipif(skip_openai, reason=reason)
|
|
@pytest.mark.asyncio
|
|
async def test_async_get_human_input():
|
|
config_list = autogen.config_list_from_json(OAI_CONFIG_LIST, KEY_LOC, filter_dict={"tags": ["gpt-3.5-turbo"]})
|
|
|
|
# create an AssistantAgent instance named "assistant"
|
|
assistant = autogen.AssistantAgent(
|
|
name="assistant",
|
|
max_consecutive_auto_reply=2,
|
|
llm_config={"config_list": config_list, "temperature": 0},
|
|
)
|
|
|
|
user_proxy = autogen.UserProxyAgent(name="user", human_input_mode="ALWAYS", code_execution_config=False)
|
|
|
|
user_proxy.a_get_human_input = AsyncMock(return_value="This is a test")
|
|
|
|
user_proxy.register_reply([autogen.Agent, None], autogen.ConversableAgent.a_check_termination_and_human_reply)
|
|
|
|
await user_proxy.a_initiate_chat(assistant, clear_history=True, message="Hello.")
|
|
# Test without message
|
|
res = await user_proxy.a_initiate_chat(assistant, clear_history=True, summary_method="reflection_with_llm")
|
|
# Assert that custom a_get_human_input was called at least once
|
|
user_proxy.a_get_human_input.assert_called()
|
|
print("Result summary:", res.summary)
|
|
print("Human input:", res.human_input)
|
|
|
|
|
|
@pytest.mark.skipif(skip_openai, reason=reason)
|
|
@pytest.mark.asyncio
|
|
async def test_async_max_turn():
|
|
config_list = autogen.config_list_from_json(OAI_CONFIG_LIST, KEY_LOC, filter_dict={"tags": ["gpt-3.5-turbo"]})
|
|
|
|
# create an AssistantAgent instance named "assistant"
|
|
assistant = autogen.AssistantAgent(
|
|
name="assistant",
|
|
max_consecutive_auto_reply=10,
|
|
llm_config={
|
|
"seed": 41,
|
|
"config_list": config_list,
|
|
},
|
|
)
|
|
|
|
user_proxy = autogen.UserProxyAgent(name="user", human_input_mode="ALWAYS", code_execution_config=False)
|
|
|
|
user_proxy.a_get_human_input = AsyncMock(return_value="Not funny. Try again.")
|
|
|
|
res = await user_proxy.a_initiate_chat(
|
|
assistant, clear_history=True, max_turns=3, message="Hello, make a joke about AI."
|
|
)
|
|
print("Result summary:", res.summary)
|
|
print("Human input:", res.human_input)
|
|
print("chat history:", res.chat_history)
|
|
assert (
|
|
len(res.chat_history) == 6
|
|
), f"Chat history should have 6 messages because max_turns is set to 3 (and user keep request try again) but has {len(res.chat_history)}."
|
|
|
|
|
|
if __name__ == "__main__":
|
|
# asyncio.run(test_async_get_human_input())
|
|
asyncio.run(test_async_max_turn())
|