langchain/tests/unit_tests/agents/test_mrkl.py
Nathan Azrak 26ec845921
Raise an exception in MKRL and Chat Output Parsers if parsing text which contains both an action and a final answer (#5609)
Raises exception if OutputParsers receive a response with both a valid
action and a final answer

Currently, if an OutputParser receives a response which includes both an
action and a final answer, they return a FinalAnswer object. This allows
the parser to accept responses which propose an action and hallucinate
an answer without the action being parsed or taken by the agent.

This PR changes the logic to:
1. store a variable checking whether a response contains the
`FINAL_ANSWER_ACTION` (this is the easier condition to check).
2. store a variable checking whether the response contains a valid
action
3. if both are present, raise a new exception stating that both are
present
4. if an action is present, return an AgentAction
5. if an answer is present, return an AgentAnswer
6. if neither is present, raise the relevant exception based around the
action format (these have been kept consistent with the prior exception
messages)

Disclaimer:
* Existing mock data included strings which did include an action and an
answer. This might indicate that prioritising returning AgentAnswer was
always correct, and I am patching out desired behaviour? @hwchase17 to
advice. Curious if there are allowed cases where this is not
hallucinating, and we do want the LLM to output an action which isn't
taken.
* I have not passed `send_to_llm` through this new exception

Fixes #5601 

## Who can review?

Community members can review the PR once tests pass. Tag
maintainers/contributors who might be interested:

@hwchase17 - project lead
@vowelparrot
2023-06-04 14:40:49 -07:00

167 lines
5.7 KiB
Python

"""Test MRKL functionality."""
from typing import Tuple
import pytest
from langchain.agents.mrkl.base import ZeroShotAgent
from langchain.agents.mrkl.output_parser import MRKLOutputParser
from langchain.agents.mrkl.prompt import FORMAT_INSTRUCTIONS, PREFIX, SUFFIX
from langchain.agents.tools import Tool
from langchain.prompts import PromptTemplate
from langchain.schema import AgentAction, OutputParserException
from tests.unit_tests.llms.fake_llm import FakeLLM
def get_action_and_input(text: str) -> Tuple[str, str]:
output = MRKLOutputParser().parse(text)
if isinstance(output, AgentAction):
return output.tool, str(output.tool_input)
else:
return "Final Answer", output.return_values["output"]
def test_get_action_and_input() -> None:
"""Test getting an action from text."""
llm_output = (
"Thought: I need to search for NBA\n" "Action: Search\n" "Action Input: NBA"
)
action, action_input = get_action_and_input(llm_output)
assert action == "Search"
assert action_input == "NBA"
def test_get_action_and_input_whitespace() -> None:
"""Test getting an action from text."""
llm_output = "Thought: I need to search for NBA\nAction: Search \nAction Input: NBA"
action, action_input = get_action_and_input(llm_output)
assert action == "Search"
assert action_input == "NBA"
def test_get_action_and_input_newline() -> None:
"""Test getting an action from text where Action Input is a code snippet."""
llm_output = (
"Now I need to write a unittest for the function.\n\n"
"Action: Python\nAction Input:\n```\nimport unittest\n\nunittest.main()\n```"
)
action, action_input = get_action_and_input(llm_output)
assert action == "Python"
assert action_input == "```\nimport unittest\n\nunittest.main()\n```"
def test_get_action_and_input_newline_after_keyword() -> None:
"""Test getting an action and action input from the text
when there is a new line before the action
(after the keywords "Action:" and "Action Input:")
"""
llm_output = """
I can use the `ls` command to list the contents of the directory \
and `grep` to search for the specific file.
Action:
Terminal
Action Input:
ls -l ~/.bashrc.d/
"""
action, action_input = get_action_and_input(llm_output)
assert action == "Terminal"
assert action_input == "ls -l ~/.bashrc.d/\n"
def test_get_action_and_input_sql_query() -> None:
"""Test getting the action and action input from the text
when the LLM output is a well formed SQL query
"""
llm_output = """
I should query for the largest single shift payment for every unique user.
Action: query_sql_db
Action Input: \
SELECT "UserName", MAX(totalpayment) FROM user_shifts GROUP BY "UserName" """
action, action_input = get_action_and_input(llm_output)
assert action == "query_sql_db"
assert (
action_input
== 'SELECT "UserName", MAX(totalpayment) FROM user_shifts GROUP BY "UserName"'
)
def test_get_final_answer() -> None:
"""Test getting final answer."""
llm_output = "Thought: I can now answer the question\n" "Final Answer: 1994"
action, action_input = get_action_and_input(llm_output)
assert action == "Final Answer"
assert action_input == "1994"
def test_get_final_answer_new_line() -> None:
"""Test getting final answer."""
llm_output = "Thought: I can now answer the question\n" "Final Answer:\n1994"
action, action_input = get_action_and_input(llm_output)
assert action == "Final Answer"
assert action_input == "1994"
def test_get_final_answer_multiline() -> None:
"""Test getting final answer that is multiline."""
llm_output = "Thought: I can now answer the question\n" "Final Answer: 1994\n1993"
action, action_input = get_action_and_input(llm_output)
assert action == "Final Answer"
assert action_input == "1994\n1993"
def test_bad_action_input_line() -> None:
"""Test handling when no action input found."""
llm_output = "Thought: I need to search for NBA\n" "Action: Search\n" "Thought: NBA"
with pytest.raises(OutputParserException) as e_info:
get_action_and_input(llm_output)
assert e_info.value.observation is not None
def test_bad_action_line() -> None:
"""Test handling when no action found."""
llm_output = (
"Thought: I need to search for NBA\n" "Thought: Search\n" "Action Input: NBA"
)
with pytest.raises(OutputParserException) as e_info:
get_action_and_input(llm_output)
assert e_info.value.observation is not None
def test_valid_action_and_answer_raises_exception() -> None:
"""Test handling when both an action and answer are found."""
llm_output = (
"Thought: I need to search for NBA\n"
"Action: Search\n"
"Action Input: NBA\n"
"Observation: founded in 1994\n"
"Thought: I can now answer the question\n"
"Final Answer: 1994"
)
with pytest.raises(OutputParserException):
get_action_and_input(llm_output)
def test_from_chains() -> None:
"""Test initializing from chains."""
chain_configs = [
Tool(name="foo", func=lambda x: "foo", description="foobar1"),
Tool(name="bar", func=lambda x: "bar", description="foobar2"),
]
agent = ZeroShotAgent.from_llm_and_tools(FakeLLM(), chain_configs)
expected_tools_prompt = "foo: foobar1\nbar: foobar2"
expected_tool_names = "foo, bar"
expected_template = "\n\n".join(
[
PREFIX,
expected_tools_prompt,
FORMAT_INSTRUCTIONS.format(tool_names=expected_tool_names),
SUFFIX,
]
)
prompt = agent.llm_chain.prompt
assert isinstance(prompt, PromptTemplate)
assert prompt.template == expected_template