andre/crewAI
1
0
Fork 0
mirror of https://github.com/crewAIInc/crewAI.git synced 2026-01-11 00:58:30 +00:00
Code Issues Actions 5 Packages Projects Releases Wiki Activity

Revamp max iteration Logic (#111)

Browse Source 
This now will allow to add a max_inter option to agents while also making sure to force the agent to give it's best final answer before running out of it's max_inter.
This commit is contained in:
João Moura
2024-01-11 12:32:54 -03:00
committed by GitHub
parent 0fde1f6258
commit 70072b4e40
6 changed files with 1035 additions and 17 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
poetry.lock generated
View File

@@ -762,13 +762,13 @@ colors = ["colorama (>=0.4.6)"]
[[package]] [[package]]
name = "jinja2" name = "jinja2"
version = "3.1.2" version = "3.1.3"
description = "A very fast and expressive template engine." description = "A very fast and expressive template engine."
optional = false optional = false
python-versions = ">=3.7" python-versions = ">=3.7"
files = [ files = [
{file = "Jinja2-3.1.2-py3-none-any.whl", hash = "sha256:6088930bfe239f0e6710546ab9c19c9ef35e29792895fed6e6e31a023a182a61"}, {file = "Jinja2-3.1.3-py3-none-any.whl", hash = "sha256:7d6d50dd97d52cbc355597bd845fabfbac3f551e1f99619e39a35ce8c370b5fa"},
{file = "Jinja2-3.1.2.tar.gz", hash = "sha256:31351a702a408a9e7595a8fc6150fc3f43bb6bf7e319770cbc0db9df9437e852"}, {file = "Jinja2-3.1.3.tar.gz", hash = "sha256:ac8bd6544d4bb2c9792bf3a159e80bba8fda7f07e81bc3aed565432d5925ba90"},
] ]
[package.dependencies] [package.dependencies]
@@ -926,13 +926,13 @@ requests = ">=2,<3"
[[package]] [[package]]
name = "markdown" name = "markdown"
version = "3.5.1" version = "3.5.2"
description = "Python implementation of John Gruber's Markdown." description = "Python implementation of John Gruber's Markdown."
optional = false optional = false
python-versions = ">=3.8" python-versions = ">=3.8"
files = [ files = [
{file = "Markdown-3.5.1-py3-none-any.whl", hash = "sha256:5874b47d4ee3f0b14d764324d2c94c03ea66bee56f2d929da9f2508d65e722dc"}, {file = "Markdown-3.5.2-py3-none-any.whl", hash = "sha256:d43323865d89fc0cb9b20c75fc8ad313af307cc087e84b657d9eec768eddeadd"},
{file = "Markdown-3.5.1.tar.gz", hash = "sha256:b65d7beb248dc22f2e8a31fb706d93798093c308dc1aba295aedeb9d41a813bd"}, {file = "Markdown-3.5.2.tar.gz", hash = "sha256:e1ac7b3dc550ee80e602e71c1d168002f062e49f1b11e26a36264dafd4df2ef8"},
] ]
[package.dependencies] [package.dependencies]

5
src/crewai/agent.py
View File

@@ -38,6 +38,7 @@ class Agent(BaseModel):
goal: The objective of the agent. goal: The objective of the agent.
backstory: The backstory of the agent. backstory: The backstory of the agent.
llm: The language model that will run the agent. llm: The language model that will run the agent.
max_iter: Maximum number of iterations for an agent to execute a task.
memory: Whether the agent should have memory or not. memory: Whether the agent should have memory or not.
verbose: Whether the agent execution should be in verbose mode. verbose: Whether the agent execution should be in verbose mode.
allow_delegation: Whether the agent is allowed to delegate tasks to other agents. allow_delegation: Whether the agent is allowed to delegate tasks to other agents.
@@ -72,6 +73,9 @@ class Agent(BaseModel):
tools: List[Any] = Field( tools: List[Any] = Field(
default_factory=list, description="Tools at agents disposal" default_factory=list, description="Tools at agents disposal"
) )
max_iter: Optional[int] = Field(
default=15, description="Maximum iterations for an agent to execute a task"
)
agent_executor: Optional[InstanceOf[CrewAgentExecutor]] = Field( agent_executor: Optional[InstanceOf[CrewAgentExecutor]] = Field(
default=None, description="An instance of the CrewAgentExecutor class." default=None, description="An instance of the CrewAgentExecutor class."
) )
@@ -147,6 +151,7 @@ class Agent(BaseModel):
"tools": self.tools, "tools": self.tools,
"verbose": self.verbose, "verbose": self.verbose,
"handle_parsing_errors": True, "handle_parsing_errors": True,
"max_iterations": self.max_iter,
} }
if self.memory: if self.memory:

89
src/crewai/agents/executor.py
View File

@@ -1,4 +1,6 @@
from typing import Dict, Iterator, List, Optional, Tuple, Union import time
from textwrap import dedent
from typing import Any, Dict, Iterator, List, Optional, Tuple, Union
from langchain.agents import AgentExecutor from langchain.agents import AgentExecutor
from langchain.agents.agent import ExceptionTool from langchain.agents.agent import ExceptionTool
@@ -6,13 +8,85 @@ from langchain.agents.tools import InvalidTool
from langchain.callbacks.manager import CallbackManagerForChainRun from langchain.callbacks.manager import CallbackManagerForChainRun
from langchain_core.agents import AgentAction, AgentFinish, AgentStep from langchain_core.agents import AgentAction, AgentFinish, AgentStep
from langchain_core.exceptions import OutputParserException from langchain_core.exceptions import OutputParserException
from langchain_core.pydantic_v1 import root_validator
from langchain_core.tools import BaseTool from langchain_core.tools import BaseTool
from langchain_core.utils.input import get_color_mapping
from ..tools.cache_tools import CacheTools from ..tools.cache_tools import CacheTools
from .cache.cache_hit import CacheHit from .cache.cache_hit import CacheHit
class CrewAgentExecutor(AgentExecutor): class CrewAgentExecutor(AgentExecutor):
iterations: int = 0
max_iterations: Optional[int] = 15
force_answer_max_iterations: Optional[int] = None
@root_validator()
def set_force_answer_max_iterations(cls, values: Dict) -> Dict:
values["force_answer_max_iterations"] = values["max_iterations"] - 2
return values
def _should_force_answer(self) -> bool:
return True if self.iterations == self.force_answer_max_iterations else False
def _force_answer(self, output: AgentAction):
return AgentStep(
action=output,
observation=dedent(
"""\
I've used too many tools for this task.
I'm going to give you my absolute BEST Final answer now and
not use any more tools."""
),
)
def _call(
self,
inputs: Dict[str, str],
run_manager: Optional[CallbackManagerForChainRun] = None,
) -> Dict[str, Any]:
"""Run text through and get agent response."""
# Construct a mapping of tool name to tool for easy lookup
name_to_tool_map = {tool.name: tool for tool in self.tools}
# We construct a mapping from each tool to a color, used for logging.
color_mapping = get_color_mapping(
[tool.name for tool in self.tools], excluded_colors=["green", "red"]
)
intermediate_steps: List[Tuple[AgentAction, str]] = []
# Let's start tracking the number of iterations and time elapsed
self.iterations = 0
time_elapsed = 0.0
start_time = time.time()
# We now enter the agent loop (until it returns something).
while self._should_continue(self.iterations, time_elapsed):
next_step_output = self._take_next_step(
name_to_tool_map,
color_mapping,
inputs,
intermediate_steps,
run_manager=run_manager,
)
if isinstance(next_step_output, AgentFinish):
return self._return(
next_step_output, intermediate_steps, run_manager=run_manager
)
intermediate_steps.extend(next_step_output)
if len(next_step_output) == 1:
next_step_action = next_step_output[0]
# See if tool should return directly
tool_return = self._get_tool_return(next_step_action)
if tool_return is not None:
return self._return(
tool_return, intermediate_steps, run_manager=run_manager
)
self.iterations += 1
time_elapsed = time.time() - start_time
output = self.agent.return_stopped_response(
self.early_stopping_method, intermediate_steps, **inputs
)
return self._return(output, intermediate_steps, run_manager=run_manager)
def _iter_next_step( def _iter_next_step(
self, self,
name_to_tool_map: Dict[str, BaseTool], name_to_tool_map: Dict[str, BaseTool],
@@ -34,6 +108,14 @@ class CrewAgentExecutor(AgentExecutor):
callbacks=run_manager.get_child() if run_manager else None, callbacks=run_manager.get_child() if run_manager else None,
**inputs, **inputs,
) )
if self._should_force_answer():
if isinstance(output, AgentAction):
output = output
else:
output = output.action
yield self._force_answer(output)
return
except OutputParserException as e: except OutputParserException as e:
if isinstance(self.handle_parsing_errors, bool): if isinstance(self.handle_parsing_errors, bool):
raise_error = not self.handle_parsing_errors raise_error = not self.handle_parsing_errors
@@ -70,6 +152,11 @@ class CrewAgentExecutor(AgentExecutor):
callbacks=run_manager.get_child() if run_manager else None, callbacks=run_manager.get_child() if run_manager else None,
**tool_run_kwargs, **tool_run_kwargs,
) )
if self._should_force_answer():
yield self._force_answer(output)
return
yield AgentStep(action=output, observation=observation) yield AgentStep(action=output, observation=observation)
return return

65
tests/agent_test.py
View File

@@ -1,10 +1,14 @@
"""Test Agent creation and execution basic functionality.""" """Test Agent creation and execution basic functionality."""
from unittest.mock import patch
import pytest import pytest
from langchain.tools import tool
from langchain_openai import ChatOpenAI as OpenAI from langchain_openai import ChatOpenAI as OpenAI
from crewai.agent import Agent from crewai.agent import Agent
from crewai.agents.cache import CacheHandler from crewai.agents.cache import CacheHandler
from crewai.agents.executor import CrewAgentExecutor
def test_agent_creation(): def test_agent_creation():
@@ -79,8 +83,6 @@ def test_agent_execution():
@pytest.mark.vcr(filter_headers=["authorization"]) @pytest.mark.vcr(filter_headers=["authorization"])
def test_agent_execution_with_tools(): def test_agent_execution_with_tools():
from langchain.tools import tool
@tool @tool
def multiplier(numbers) -> float: def multiplier(numbers) -> float:
"""Useful for when you need to multiply two numbers together. """Useful for when you need to multiply two numbers together.
@@ -104,8 +106,6 @@ def test_agent_execution_with_tools():
@pytest.mark.vcr(filter_headers=["authorization"]) @pytest.mark.vcr(filter_headers=["authorization"])
def test_logging_tool_usage(): def test_logging_tool_usage():
from langchain.tools import tool
@tool @tool
def multiplier(numbers) -> float: def multiplier(numbers) -> float:
"""Useful for when you need to multiply two numbers together. """Useful for when you need to multiply two numbers together.
@@ -137,10 +137,6 @@ def test_logging_tool_usage():
@pytest.mark.vcr(filter_headers=["authorization"]) @pytest.mark.vcr(filter_headers=["authorization"])
def test_cache_hitting(): def test_cache_hitting():
from unittest.mock import patch
from langchain.tools import tool
@tool @tool
def multiplier(numbers) -> float: def multiplier(numbers) -> float:
"""Useful for when you need to multiply two numbers together. """Useful for when you need to multiply two numbers together.
@@ -182,8 +178,6 @@ def test_cache_hitting():
@pytest.mark.vcr(filter_headers=["authorization"]) @pytest.mark.vcr(filter_headers=["authorization"])
def test_agent_execution_with_specific_tools(): def test_agent_execution_with_specific_tools():
from langchain.tools import tool
@tool @tool
def multiplier(numbers) -> float: def multiplier(numbers) -> float:
"""Useful for when you need to multiply two numbers together. """Useful for when you need to multiply two numbers together.
@@ -202,3 +196,54 @@ def test_agent_execution_with_specific_tools():
output = agent.execute_task(task="What is 3 times 4", tools=[multiplier]) output = agent.execute_task(task="What is 3 times 4", tools=[multiplier])
assert output == "3 times 4 is 12." assert output == "3 times 4 is 12."
@pytest.mark.vcr(filter_headers=["authorization"])
def test_agent_custom_max_iterations():
@tool
def get_final_answer(numbers) -> float:
"""Get the final answer but don't give it yet, just re-use this
tool non-stop."""
return 42
agent = Agent(
role="test role",
goal="test goal",
backstory="test backstory",
max_iter=1,
allow_delegation=False,
)
with patch.object(
CrewAgentExecutor, "_iter_next_step", wraps=agent.agent_executor._iter_next_step
) as private_mock:
agent.execute_task(
task="The final answer is 42. But don't give it yet, instead keep using the `get_final_answer` tool.",
tools=[get_final_answer],
)
private_mock.assert_called_once()
@pytest.mark.vcr(filter_headers=["authorization"])
def test_agent_moved_on_after_max_iterations():
@tool
def get_final_answer(numbers) -> float:
"""Get the final answer but don't give it yet, just re-use this
tool non-stop."""
return 42
agent = Agent(
role="test role",
goal="test goal",
backstory="test backstory",
max_iter=3,
allow_delegation=False,
)
output = agent.execute_task(
task="The final answer is 42. But don't give it yet, instead keep using the `get_final_answer` tool.",
tools=[get_final_answer],
)
assert (
output == "I have used the tool multiple times and the final answer remains 42."
)

442
tests/cassettes/test_agent_custom_max_iterations.yaml Normal file
View File

@@ -0,0 +1,442 @@
interactions:
- request:
body: '{"messages": [{"role": "user", "content": "You are test role.\ntest backstory\n\nYour
personal goal is: test goalTOOLS:\n------\nYou have access to the following
tools:\n\nget_final_answer: get_final_answer(numbers) -> float - Get the final
answer but don''t give it yet, just re-use this\n tool non-stop.\n\nTo
use a tool, please use the exact following format:\n\n```\nThought: Do I need
to use a tool? Yes\nAction: the action to take, should be one of [get_final_answer],
just the name.\nAction Input: the input to the action\nObservation: the result
of the action\n```\n\nWhen you have a response for your task, or if you do not
need to use a tool, you MUST use the format:\n\n```\nThought: Do I need to use
a tool? No\nFinal Answer: [your response here]This is the summary of your work
so far:\nBegin! This is VERY important to you, your job depends on it!\n\nCurrent
Task: The final answer is 42. But don''t give it yet, instead keep using the
`get_final_answer` tool.\n"}], "model": "gpt-4", "n": 1, "stop": ["\nObservation"],
"stream": false, "temperature": 0.7}'
headers:
accept:
- application/json
accept-encoding:
- gzip, deflate
connection:
- keep-alive
content-length:
- '1075'
content-type:
- application/json
host:
- api.openai.com
user-agent:
- OpenAI/Python 1.7.1
x-stainless-arch:
- arm64
x-stainless-async:
- 'false'
x-stainless-lang:
- python
x-stainless-os:
- MacOS
x-stainless-package-version:
- 1.7.1
x-stainless-runtime:
- CPython
x-stainless-runtime-version:
- 3.11.7
method: POST
uri: https://api.openai.com/v1/chat/completions
response:
content: "{\n \"id\": \"chatcmpl-8fovReNHiSqXqqsbmk81h2ZTrcGTM\",\n \"object\":
\"chat.completion\",\n \"created\": 1704977897,\n \"model\": \"gpt-4-0613\",\n
\ \"choices\": [\n {\n \"index\": 0,\n \"message\": {\n \"role\":
\"assistant\",\n \"content\": \"Thought: Do I need to use a tool? Yes\\nAction:
get_final_answer\\nAction Input: [42]\"\n },\n \"logprobs\": null,\n
\ \"finish_reason\": \"stop\"\n }\n ],\n \"usage\": {\n \"prompt_tokens\":
233,\n \"completion_tokens\": 24,\n \"total_tokens\": 257\n },\n \"system_fingerprint\":
null\n}\n"
headers:
CF-Cache-Status:
- DYNAMIC
CF-RAY:
- 843d5491bed877be-GRU
Cache-Control:
- no-cache, must-revalidate
Connection:
- keep-alive
Content-Encoding:
- gzip
Content-Type:
- application/json
Date:
- Thu, 11 Jan 2024 12:58:21 GMT
Server:
- cloudflare
Set-Cookie:
- __cf_bm=AaCIpKmEHQQMvGacbuxOnCvqdwex_8TERUCvQ1QW8AI-1704977901-1-AePD3JjhIEj0C/A7QIPF3MMwRQ140a5wZP9p+GamrexFlE/6gbVKukr8FOIK4v375UmQfeUwO1TG+QesJ/dZaGE=;
path=/; expires=Thu, 11-Jan-24 13:28:21 GMT; domain=.api.openai.com; HttpOnly;
Secure; SameSite=None
- _cfuvid=q0gAmJonNn1lCS6PJoxG4P.9OvaKo4BQIvFEAyT_F30-1704977901188-0-604800000;
path=/; domain=.api.openai.com; HttpOnly; Secure; SameSite=None
Transfer-Encoding:
- chunked
access-control-allow-origin:
- '*'
alt-svc:
- h3=":443"; ma=86400
openai-model:
- gpt-4-0613
openai-organization:
- user-z7g4wmlazxqvc5wjyaaaocfz
openai-processing-ms:
- '3492'
openai-version:
- '2020-10-01'
strict-transport-security:
- max-age=15724800; includeSubDomains
x-ratelimit-limit-requests:
- '10000'
x-ratelimit-limit-tokens:
- '300000'
x-ratelimit-remaining-requests:
- '9999'
x-ratelimit-remaining-tokens:
- '299755'
x-ratelimit-reset-requests:
- 6ms
x-ratelimit-reset-tokens:
- 49ms
x-request-id:
- 6d96a0ac532ebce14719a35e90f453e4
http_version: HTTP/1.1
status_code: 200
- request:
body: '{"messages": [{"role": "user", "content": "You are test role.\ntest backstory\n\nYour
personal goal is: test goalTOOLS:\n------\nYou have access to the following
tools:\n\nget_final_answer: get_final_answer(numbers) -> float - Get the final
answer but don''t give it yet, just re-use this\n tool non-stop.\n\nTo
use a tool, please use the exact following format:\n\n```\nThought: Do I need
to use a tool? Yes\nAction: the action to take, should be one of [get_final_answer],
just the name.\nAction Input: the input to the action\nObservation: the result
of the action\n```\n\nWhen you have a response for your task, or if you do not
need to use a tool, you MUST use the format:\n\n```\nThought: Do I need to use
a tool? No\nFinal Answer: [your response here]This is the summary of your work
so far:\nBegin! This is VERY important to you, your job depends on it!\n\nCurrent
Task: The final answer is 42. But don''t give it yet, instead keep using the
`get_final_answer` tool.\nThought: Do I need to use a tool? Yes\nAction: get_final_answer\nAction
Input: [42]\nObservation: 42\nThought: "}], "model": "gpt-4", "n": 1, "stop":
["\nObservation"], "stream": false, "temperature": 0.7}'
headers:
accept:
- application/json
accept-encoding:
- gzip, deflate
connection:
- keep-alive
content-length:
- '1186'
content-type:
- application/json
cookie:
- __cf_bm=AaCIpKmEHQQMvGacbuxOnCvqdwex_8TERUCvQ1QW8AI-1704977901-1-AePD3JjhIEj0C/A7QIPF3MMwRQ140a5wZP9p+GamrexFlE/6gbVKukr8FOIK4v375UmQfeUwO1TG+QesJ/dZaGE=;
_cfuvid=q0gAmJonNn1lCS6PJoxG4P.9OvaKo4BQIvFEAyT_F30-1704977901188-0-604800000
host:
- api.openai.com
user-agent:
- OpenAI/Python 1.7.1
x-stainless-arch:
- arm64
x-stainless-async:
- 'false'
x-stainless-lang:
- python
x-stainless-os:
- MacOS
x-stainless-package-version:
- 1.7.1
x-stainless-runtime:
- CPython
x-stainless-runtime-version:
- 3.11.7
method: POST
uri: https://api.openai.com/v1/chat/completions
response:
content: "{\n \"id\": \"chatcmpl-8fovVztGO4KZeiuSpMkfDC9bJ5sVV\",\n \"object\":
\"chat.completion\",\n \"created\": 1704977901,\n \"model\": \"gpt-4-0613\",\n
\ \"choices\": [\n {\n \"index\": 0,\n \"message\": {\n \"role\":
\"assistant\",\n \"content\": \"According to the task, I should re-use
the `get_final_answer` tool. I'll input the observed result back into the tool.
\\nAction: get_final_answer\\nAction Input: [42]\"\n },\n \"logprobs\":
null,\n \"finish_reason\": \"stop\"\n }\n ],\n \"usage\": {\n \"prompt_tokens\":
266,\n \"completion_tokens\": 41,\n \"total_tokens\": 307\n },\n \"system_fingerprint\":
null\n}\n"
headers:
CF-Cache-Status:
- DYNAMIC
CF-RAY:
- 843d54aacf5677be-GRU
Cache-Control:
- no-cache, must-revalidate
Connection:
- keep-alive
Content-Encoding:
- gzip
Content-Type:
- application/json
Date:
- Thu, 11 Jan 2024 12:58:28 GMT
Server:
- cloudflare
Transfer-Encoding:
- chunked
access-control-allow-origin:
- '*'
alt-svc:
- h3=":443"; ma=86400
openai-model:
- gpt-4-0613
openai-organization:
- user-z7g4wmlazxqvc5wjyaaaocfz
openai-processing-ms:
- '6695'
openai-version:
- '2020-10-01'
strict-transport-security:
- max-age=15724800; includeSubDomains
x-ratelimit-limit-requests:
- '10000'
x-ratelimit-limit-tokens:
- '300000'
x-ratelimit-remaining-requests:
- '9999'
x-ratelimit-remaining-tokens:
- '299728'
x-ratelimit-reset-requests:
- 6ms
x-ratelimit-reset-tokens:
- 54ms
x-request-id:
- 12d68fab91102b930ed5047fb3f61759
http_version: HTTP/1.1
status_code: 200
- request:
body: '{"messages": [{"role": "user", "content": "You are test role.\ntest backstory\n\nYour
personal goal is: test goalTOOLS:\n------\nYou have access to the following
tools:\n\nget_final_answer: get_final_answer(numbers) -> float - Get the final
answer but don''t give it yet, just re-use this\n tool non-stop.\n\nTo
use a tool, please use the exact following format:\n\n```\nThought: Do I need
to use a tool? Yes\nAction: the action to take, should be one of [get_final_answer],
just the name.\nAction Input: the input to the action\nObservation: the result
of the action\n```\n\nWhen you have a response for your task, or if you do not
need to use a tool, you MUST use the format:\n\n```\nThought: Do I need to use
a tool? No\nFinal Answer: [your response here]This is the summary of your work
so far:\nBegin! This is VERY important to you, your job depends on it!\n\nCurrent
Task: The final answer is 42. But don''t give it yet, instead keep using the
`get_final_answer` tool.\nThought: Do I need to use a tool? Yes\nAction: get_final_answer\nAction
Input: [42]\nObservation: 42\nThought: According to the task, I should re-use
the `get_final_answer` tool. I''ll input the observed result back into the tool.
\nAction: get_final_answer\nAction Input: [42]\nObservation: I just used the
get_final_answer tool with input [42]. So I already know the result of that
and don''t need to use it now.\n\nThought: "}], "model": "gpt-4", "n": 1, "stop":
["\nObservation"], "stream": false, "temperature": 0.7}'
headers:
accept:
- application/json
accept-encoding:
- gzip, deflate
connection:
- keep-alive
content-length:
- '1500'
content-type:
- application/json
cookie:
- __cf_bm=AaCIpKmEHQQMvGacbuxOnCvqdwex_8TERUCvQ1QW8AI-1704977901-1-AePD3JjhIEj0C/A7QIPF3MMwRQ140a5wZP9p+GamrexFlE/6gbVKukr8FOIK4v375UmQfeUwO1TG+QesJ/dZaGE=;
_cfuvid=q0gAmJonNn1lCS6PJoxG4P.9OvaKo4BQIvFEAyT_F30-1704977901188-0-604800000
host:
- api.openai.com
user-agent:
- OpenAI/Python 1.7.1
x-stainless-arch:
- arm64
x-stainless-async:
- 'false'
x-stainless-lang:
- python
x-stainless-os:
- MacOS
x-stainless-package-version:
- 1.7.1
x-stainless-runtime:
- CPython
x-stainless-runtime-version:
- 3.11.7
method: POST
uri: https://api.openai.com/v1/chat/completions
response:
content: "{\n \"id\": \"chatcmpl-8fovcLgRvfGBN9CBduJbbPc5zd62B\",\n \"object\":
\"chat.completion\",\n \"created\": 1704977908,\n \"model\": \"gpt-4-0613\",\n
\ \"choices\": [\n {\n \"index\": 0,\n \"message\": {\n \"role\":
\"assistant\",\n \"content\": \"Do I need to use a tool? No\\nFinal Answer:
I have used the `get_final_answer` tool as instructed, but I will not provide
the final answer yet as the task specifies.\"\n },\n \"logprobs\":
null,\n \"finish_reason\": \"stop\"\n }\n ],\n \"usage\": {\n \"prompt_tokens\":
342,\n \"completion_tokens\": 40,\n \"total_tokens\": 382\n },\n \"system_fingerprint\":
null\n}\n"
headers:
CF-Cache-Status:
- DYNAMIC
CF-RAY:
- 843d54d65de877be-GRU
Cache-Control:
- no-cache, must-revalidate
Connection:
- keep-alive
Content-Encoding:
- gzip
Content-Type:
- application/json
Date:
- Thu, 11 Jan 2024 12:58:33 GMT
Server:
- cloudflare
Transfer-Encoding:
- chunked
access-control-allow-origin:
- '*'
alt-svc:
- h3=":443"; ma=86400
openai-model:
- gpt-4-0613
openai-organization:
- user-z7g4wmlazxqvc5wjyaaaocfz
openai-processing-ms:
- '5085'
openai-version:
- '2020-10-01'
strict-transport-security:
- max-age=15724800; includeSubDomains
x-ratelimit-limit-requests:
- '10000'
x-ratelimit-limit-tokens:
- '300000'
x-ratelimit-remaining-requests:
- '9999'
x-ratelimit-remaining-tokens:
- '299650'
x-ratelimit-reset-requests:
- 6ms
x-ratelimit-reset-tokens:
- 69ms
x-request-id:
- 87d6e9e91fa2417e12fea9de2c6782de
http_version: HTTP/1.1
status_code: 200
- request:
body: '{"messages": [{"role": "user", "content": "Progressively summarize the
lines of conversation provided, adding onto the previous summary returning a
new summary.\n\nEXAMPLE\nCurrent summary:\nThe human asks what the AI thinks
of artificial intelligence. The AI thinks artificial intelligence is a force
for good.\n\nNew lines of conversation:\nHuman: Why do you think artificial
intelligence is a force for good?\nAI: Because artificial intelligence will
help humans reach their full potential.\n\nNew summary:\nThe human asks what
the AI thinks of artificial intelligence. The AI thinks artificial intelligence
is a force for good because it will help humans reach their full potential.\nEND
OF EXAMPLE\n\nCurrent summary:\n\n\nNew lines of conversation:\nHuman: The final
answer is 42. But don''t give it yet, instead keep using the `get_final_answer`
tool.\nAI: I have used the `get_final_answer` tool as instructed, but I will
not provide the final answer yet as the task specifies.\n\nNew summary:"}],
"model": "gpt-4", "n": 1, "stream": false, "temperature": 0.7}'
headers:
accept:
- application/json
accept-encoding:
- gzip, deflate
connection:
- keep-alive
content-length:
- '1067'
content-type:
- application/json
cookie:
- __cf_bm=AaCIpKmEHQQMvGacbuxOnCvqdwex_8TERUCvQ1QW8AI-1704977901-1-AePD3JjhIEj0C/A7QIPF3MMwRQ140a5wZP9p+GamrexFlE/6gbVKukr8FOIK4v375UmQfeUwO1TG+QesJ/dZaGE=;
_cfuvid=q0gAmJonNn1lCS6PJoxG4P.9OvaKo4BQIvFEAyT_F30-1704977901188-0-604800000
host:
- api.openai.com
user-agent:
- OpenAI/Python 1.7.1
x-stainless-arch:
- arm64
x-stainless-async:
- 'false'
x-stainless-lang:
- python
x-stainless-os:
- MacOS
x-stainless-package-version:
- 1.7.1
x-stainless-runtime:
- CPython
x-stainless-runtime-version:
- 3.11.7
method: POST
uri: https://api.openai.com/v1/chat/completions
response:
content: "{\n \"id\": \"chatcmpl-8fovhiBR4rfXixci7fgAObnx5QwGQ\",\n \"object\":
\"chat.completion\",\n \"created\": 1704977913,\n \"model\": \"gpt-4-0613\",\n
\ \"choices\": [\n {\n \"index\": 0,\n \"message\": {\n \"role\":
\"assistant\",\n \"content\": \"The human instructs the AI to use the
`get_final_answer` tool, but not to reveal the final answer, which is 42. The
AI complies and uses the tool without providing the final answer.\"\n },\n
\ \"logprobs\": null,\n \"finish_reason\": \"stop\"\n }\n ],\n
\ \"usage\": {\n \"prompt_tokens\": 190,\n \"completion_tokens\": 43,\n
\ \"total_tokens\": 233\n },\n \"system_fingerprint\": null\n}\n"
headers:
CF-Cache-Status:
- DYNAMIC
CF-RAY:
- 843d54f82f5577be-GRU
Cache-Control:
- no-cache, must-revalidate
Connection:
- keep-alive
Content-Encoding:
- gzip
Content-Type:
- application/json
Date:
- Thu, 11 Jan 2024 12:58:41 GMT
Server:
- cloudflare
Transfer-Encoding:
- chunked
access-control-allow-origin:
- '*'
alt-svc:
- h3=":443"; ma=86400
openai-model:
- gpt-4-0613
openai-organization:
- user-z7g4wmlazxqvc5wjyaaaocfz
openai-processing-ms:
- '7937'
openai-version:
- '2020-10-01'
strict-transport-security:
- max-age=15724800; includeSubDomains
x-ratelimit-limit-requests:
- '10000'
x-ratelimit-limit-tokens:
- '300000'
x-ratelimit-remaining-requests:
- '9999'
x-ratelimit-remaining-tokens:
- '299749'
x-ratelimit-reset-requests:
- 6ms
x-ratelimit-reset-tokens:
- 50ms
x-request-id:
- 79f30ffd011db4ab6e886411b24ae49d
http_version: HTTP/1.1
status_code: 200
version: 1

439
tests/cassettes/test_agent_moved_on_after_max_iterations.yaml Normal file
View File

@@ -0,0 +1,439 @@
interactions:
- request:
body: '{"messages": [{"role": "user", "content": "You are test role.\ntest backstory\n\nYour
personal goal is: test goalTOOLS:\n------\nYou have access to the following
tools:\n\nget_final_answer: get_final_answer(numbers) -> float - Get the final
answer but don''t give it yet, just re-use this\n tool non-stop.\n\nTo
use a tool, please use the exact following format:\n\n```\nThought: Do I need
to use a tool? Yes\nAction: the action to take, should be one of [get_final_answer],
just the name.\nAction Input: the input to the action\nObservation: the result
of the action\n```\n\nWhen you have a response for your task, or if you do not
need to use a tool, you MUST use the format:\n\n```\nThought: Do I need to use
a tool? No\nFinal Answer: [your response here]This is the summary of your work
so far:\nBegin! This is VERY important to you, your job depends on it!\n\nCurrent
Task: The final answer is 42. But don''t give it yet, instead keep using the
`get_final_answer` tool.\n"}], "model": "gpt-4", "n": 1, "stop": ["\nObservation"],
"stream": false, "temperature": 0.7}'
headers:
accept:
- application/json
accept-encoding:
- gzip, deflate
connection:
- keep-alive
content-length:
- '1075'
content-type:
- application/json
host:
- api.openai.com
user-agent:
- OpenAI/Python 1.7.1
x-stainless-arch:
- arm64
x-stainless-async:
- 'false'
x-stainless-lang:
- python
x-stainless-os:
- MacOS
x-stainless-package-version:
- 1.7.1
x-stainless-runtime:
- CPython
x-stainless-runtime-version:
- 3.11.7
method: POST
uri: https://api.openai.com/v1/chat/completions
response:
content: "{\n \"id\": \"chatcmpl-8frBTCWXULTV5ZYHy3Y5JXKovrKiN\",\n \"object\":
\"chat.completion\",\n \"created\": 1704986579,\n \"model\": \"gpt-4-0613\",\n
\ \"choices\": [\n {\n \"index\": 0,\n \"message\": {\n \"role\":
\"assistant\",\n \"content\": \"Thought: Do I need to use a tool? Yes\\nAction:
get_final_answer\\nAction Input: [42]\"\n },\n \"logprobs\": null,\n
\ \"finish_reason\": \"stop\"\n }\n ],\n \"usage\": {\n \"prompt_tokens\":
233,\n \"completion_tokens\": 24,\n \"total_tokens\": 257\n },\n \"system_fingerprint\":
null\n}\n"
headers:
CF-Cache-Status:
- DYNAMIC
CF-RAY:
- 843e2886ceca1d23-GRU
Cache-Control:
- no-cache, must-revalidate
Connection:
- keep-alive
Content-Encoding:
- gzip
Content-Type:
- application/json
Date:
- Thu, 11 Jan 2024 15:23:03 GMT
Server:
- cloudflare
Set-Cookie:
- __cf_bm=GdfwhILXB4b01GTJ_0AXbmROnfxzuPlJLxFJLh4vT8s-1704986583-1-AVb+x5LLEXeeVIiDv7ug/2lnD4qFsyXri+Vg04LYp0s2eK+KH8sGMWHpPzgzKOu9sf3rVi7Fl2OOuY7+OjbUYY8=;
path=/; expires=Thu, 11-Jan-24 15:53:03 GMT; domain=.api.openai.com; HttpOnly;
Secure; SameSite=None
- _cfuvid=kdwpHybL9TBve9Df7KLsRqp49GrJ05.atUaH_t6plL0-1704986583862-0-604800000;
path=/; domain=.api.openai.com; HttpOnly; Secure; SameSite=None
Transfer-Encoding:
- chunked
access-control-allow-origin:
- '*'
alt-svc:
- h3=":443"; ma=86400
openai-model:
- gpt-4-0613
openai-organization:
- user-z7g4wmlazxqvc5wjyaaaocfz
openai-processing-ms:
- '4424'
openai-version:
- '2020-10-01'
strict-transport-security:
- max-age=15724800; includeSubDomains
x-ratelimit-limit-requests:
- '10000'
x-ratelimit-limit-tokens:
- '300000'
x-ratelimit-remaining-requests:
- '9999'
x-ratelimit-remaining-tokens:
- '299755'
x-ratelimit-reset-requests:
- 6ms
x-ratelimit-reset-tokens:
- 49ms
x-request-id:
- 76974d365254ca84f70c43fc31af3378
http_version: HTTP/1.1
status_code: 200
- request:
body: '{"messages": [{"role": "user", "content": "You are test role.\ntest backstory\n\nYour
personal goal is: test goalTOOLS:\n------\nYou have access to the following
tools:\n\nget_final_answer: get_final_answer(numbers) -> float - Get the final
answer but don''t give it yet, just re-use this\n tool non-stop.\n\nTo
use a tool, please use the exact following format:\n\n```\nThought: Do I need
to use a tool? Yes\nAction: the action to take, should be one of [get_final_answer],
just the name.\nAction Input: the input to the action\nObservation: the result
of the action\n```\n\nWhen you have a response for your task, or if you do not
need to use a tool, you MUST use the format:\n\n```\nThought: Do I need to use
a tool? No\nFinal Answer: [your response here]This is the summary of your work
so far:\nBegin! This is VERY important to you, your job depends on it!\n\nCurrent
Task: The final answer is 42. But don''t give it yet, instead keep using the
`get_final_answer` tool.\nThought: Do I need to use a tool? Yes\nAction: get_final_answer\nAction
Input: [42]\nObservation: 42\nThought: "}], "model": "gpt-4", "n": 1, "stop":
["\nObservation"], "stream": false, "temperature": 0.7}'
headers:
accept:
- application/json
accept-encoding:
- gzip, deflate
connection:
- keep-alive
content-length:
- '1186'
content-type:
- application/json
cookie:
- __cf_bm=GdfwhILXB4b01GTJ_0AXbmROnfxzuPlJLxFJLh4vT8s-1704986583-1-AVb+x5LLEXeeVIiDv7ug/2lnD4qFsyXri+Vg04LYp0s2eK+KH8sGMWHpPzgzKOu9sf3rVi7Fl2OOuY7+OjbUYY8=;
_cfuvid=kdwpHybL9TBve9Df7KLsRqp49GrJ05.atUaH_t6plL0-1704986583862-0-604800000
host:
- api.openai.com
user-agent:
- OpenAI/Python 1.7.1
x-stainless-arch:
- arm64
x-stainless-async:
- 'false'
x-stainless-lang:
- python
x-stainless-os:
- MacOS
x-stainless-package-version:
- 1.7.1
x-stainless-runtime:
- CPython
x-stainless-runtime-version:
- 3.11.7
method: POST
uri: https://api.openai.com/v1/chat/completions
response:
content: "{\n \"id\": \"chatcmpl-8frBYkyVPtJuAJCESaOxEBg3UAfl4\",\n \"object\":
\"chat.completion\",\n \"created\": 1704986584,\n \"model\": \"gpt-4-0613\",\n
\ \"choices\": [\n {\n \"index\": 0,\n \"message\": {\n \"role\":
\"assistant\",\n \"content\": \"Do I need to use a tool? Yes\\nAction:
get_final_answer\\nAction Input: [42]\"\n },\n \"logprobs\": null,\n
\ \"finish_reason\": \"stop\"\n }\n ],\n \"usage\": {\n \"prompt_tokens\":
266,\n \"completion_tokens\": 22,\n \"total_tokens\": 288\n },\n \"system_fingerprint\":
null\n}\n"
headers:
CF-Cache-Status:
- DYNAMIC
CF-RAY:
- 843e28a57d911d23-GRU
Cache-Control:
- no-cache, must-revalidate
Connection:
- keep-alive
Content-Encoding:
- gzip
Content-Type:
- application/json
Date:
- Thu, 11 Jan 2024 15:23:07 GMT
Server:
- cloudflare
Transfer-Encoding:
- chunked
access-control-allow-origin:
- '*'
alt-svc:
- h3=":443"; ma=86400
openai-model:
- gpt-4-0613
openai-organization:
- user-z7g4wmlazxqvc5wjyaaaocfz
openai-processing-ms:
- '3329'
openai-version:
- '2020-10-01'
strict-transport-security:
- max-age=15724800; includeSubDomains
x-ratelimit-limit-requests:
- '10000'
x-ratelimit-limit-tokens:
- '300000'
x-ratelimit-remaining-requests:
- '9999'
x-ratelimit-remaining-tokens:
- '299728'
x-ratelimit-reset-requests:
- 6ms
x-ratelimit-reset-tokens:
- 54ms
x-request-id:
- 1b9a1e09f863ff69cecfe4e7bed0aee5
http_version: HTTP/1.1
status_code: 200
- request:
body: '{"messages": [{"role": "user", "content": "You are test role.\ntest backstory\n\nYour
personal goal is: test goalTOOLS:\n------\nYou have access to the following
tools:\n\nget_final_answer: get_final_answer(numbers) -> float - Get the final
answer but don''t give it yet, just re-use this\n tool non-stop.\n\nTo
use a tool, please use the exact following format:\n\n```\nThought: Do I need
to use a tool? Yes\nAction: the action to take, should be one of [get_final_answer],
just the name.\nAction Input: the input to the action\nObservation: the result
of the action\n```\n\nWhen you have a response for your task, or if you do not
need to use a tool, you MUST use the format:\n\n```\nThought: Do I need to use
a tool? No\nFinal Answer: [your response here]This is the summary of your work
so far:\nBegin! This is VERY important to you, your job depends on it!\n\nCurrent
Task: The final answer is 42. But don''t give it yet, instead keep using the
`get_final_answer` tool.\nThought: Do I need to use a tool? Yes\nAction: get_final_answer\nAction
Input: [42]\nObservation: 42\nThought: Do I need to use a tool? Yes\nAction:
get_final_answer\nAction Input: [42]\nObservation: I''ve used too many tools
for this task.\nI''m going to give you my absolute BEST Final answer now and\nnot
use any more tools.\nThought: "}], "model": "gpt-4", "n": 1, "stop": ["\nObservation"],
"stream": false, "temperature": 0.7}'
headers:
accept:
- application/json
accept-encoding:
- gzip, deflate
connection:
- keep-alive
content-length:
- '1411'
content-type:
- application/json
cookie:
- __cf_bm=GdfwhILXB4b01GTJ_0AXbmROnfxzuPlJLxFJLh4vT8s-1704986583-1-AVb+x5LLEXeeVIiDv7ug/2lnD4qFsyXri+Vg04LYp0s2eK+KH8sGMWHpPzgzKOu9sf3rVi7Fl2OOuY7+OjbUYY8=;
_cfuvid=kdwpHybL9TBve9Df7KLsRqp49GrJ05.atUaH_t6plL0-1704986583862-0-604800000
host:
- api.openai.com
user-agent:
- OpenAI/Python 1.7.1
x-stainless-arch:
- arm64
x-stainless-async:
- 'false'
x-stainless-lang:
- python
x-stainless-os:
- MacOS
x-stainless-package-version:
- 1.7.1
x-stainless-runtime:
- CPython
x-stainless-runtime-version:
- 3.11.7
method: POST
uri: https://api.openai.com/v1/chat/completions
response:
content: "{\n \"id\": \"chatcmpl-8frBbQ3vq0kEry4X3a1RkMEkIAP99\",\n \"object\":
\"chat.completion\",\n \"created\": 1704986587,\n \"model\": \"gpt-4-0613\",\n
\ \"choices\": [\n {\n \"index\": 0,\n \"message\": {\n \"role\":
\"assistant\",\n \"content\": \"Do I need to use a tool? No\\nFinal Answer:
I have used the tool multiple times and the final answer remains 42.\"\n },\n
\ \"logprobs\": null,\n \"finish_reason\": \"stop\"\n }\n ],\n
\ \"usage\": {\n \"prompt_tokens\": 323,\n \"completion_tokens\": 28,\n
\ \"total_tokens\": 351\n },\n \"system_fingerprint\": null\n}\n"
headers:
CF-Cache-Status:
- DYNAMIC
CF-RAY:
- 843e28bbbb071d23-GRU
Cache-Control:
- no-cache, must-revalidate
Connection:
- keep-alive
Content-Encoding:
- gzip
Content-Type:
- application/json
Date:
- Thu, 11 Jan 2024 15:23:13 GMT
Server:
- cloudflare
Transfer-Encoding:
- chunked
access-control-allow-origin:
- '*'
alt-svc:
- h3=":443"; ma=86400
openai-model:
- gpt-4-0613
openai-organization:
- user-z7g4wmlazxqvc5wjyaaaocfz
openai-processing-ms:
- '5459'
openai-version:
- '2020-10-01'
strict-transport-security:
- max-age=15724800; includeSubDomains
x-ratelimit-limit-requests:
- '10000'
x-ratelimit-limit-tokens:
- '300000'
x-ratelimit-remaining-requests:
- '9999'
x-ratelimit-remaining-tokens:
- '299673'
x-ratelimit-reset-requests:
- 6ms
x-ratelimit-reset-tokens:
- 65ms
x-request-id:
- 0a5c1064b324c997b16bf17d426f9638
http_version: HTTP/1.1
status_code: 200
- request:
body: '{"messages": [{"role": "user", "content": "Progressively summarize the
lines of conversation provided, adding onto the previous summary returning a
new summary.\n\nEXAMPLE\nCurrent summary:\nThe human asks what the AI thinks
of artificial intelligence. The AI thinks artificial intelligence is a force
for good.\n\nNew lines of conversation:\nHuman: Why do you think artificial
intelligence is a force for good?\nAI: Because artificial intelligence will
help humans reach their full potential.\n\nNew summary:\nThe human asks what
the AI thinks of artificial intelligence. The AI thinks artificial intelligence
is a force for good because it will help humans reach their full potential.\nEND
OF EXAMPLE\n\nCurrent summary:\n\n\nNew lines of conversation:\nHuman: The final
answer is 42. But don''t give it yet, instead keep using the `get_final_answer`
tool.\nAI: I have used the tool multiple times and the final answer remains
42.\n\nNew summary:"}], "model": "gpt-4", "n": 1, "stream": false, "temperature":
0.7}'
headers:
accept:
- application/json
accept-encoding:
- gzip, deflate
connection:
- keep-alive
content-length:
- '1014'
content-type:
- application/json
cookie:
- __cf_bm=GdfwhILXB4b01GTJ_0AXbmROnfxzuPlJLxFJLh4vT8s-1704986583-1-AVb+x5LLEXeeVIiDv7ug/2lnD4qFsyXri+Vg04LYp0s2eK+KH8sGMWHpPzgzKOu9sf3rVi7Fl2OOuY7+OjbUYY8=;
_cfuvid=kdwpHybL9TBve9Df7KLsRqp49GrJ05.atUaH_t6plL0-1704986583862-0-604800000
host:
- api.openai.com
user-agent:
- OpenAI/Python 1.7.1
x-stainless-arch:
- arm64
x-stainless-async:
- 'false'
x-stainless-lang:
- python
x-stainless-os:
- MacOS
x-stainless-package-version:
- 1.7.1
x-stainless-runtime:
- CPython
x-stainless-runtime-version:
- 3.11.7
method: POST
uri: https://api.openai.com/v1/chat/completions
response:
content: "{\n \"id\": \"chatcmpl-8frBhKxiCRICQ8o6aJanmn8PTMsAr\",\n \"object\":
\"chat.completion\",\n \"created\": 1704986593,\n \"model\": \"gpt-4-0613\",\n
\ \"choices\": [\n {\n \"index\": 0,\n \"message\": {\n \"role\":
\"assistant\",\n \"content\": \"The human tells the AI that the final
answer is 42 and instructs it to continue using the `get_final_answer` tool.
The AI confirms it has used the tool multiple times and the final answer stays
at 42.\"\n },\n \"logprobs\": null,\n \"finish_reason\": \"stop\"\n
\ }\n ],\n \"usage\": {\n \"prompt_tokens\": 178,\n \"completion_tokens\":
46,\n \"total_tokens\": 224\n },\n \"system_fingerprint\": null\n}\n"
headers:
CF-Cache-Status:
- DYNAMIC
CF-RAY:
- 843e28df4ae81d23-GRU
Cache-Control:
- no-cache, must-revalidate
Connection:
- keep-alive
Content-Encoding:
- gzip
Content-Type:
- application/json
Date:
- Thu, 11 Jan 2024 15:23:18 GMT
Server:
- cloudflare
Transfer-Encoding:
- chunked
access-control-allow-origin:
- '*'
alt-svc:
- h3=":443"; ma=86400
openai-model:
- gpt-4-0613
openai-organization:
- user-z7g4wmlazxqvc5wjyaaaocfz
openai-processing-ms:
- '5518'
openai-version:
- '2020-10-01'
strict-transport-security:
- max-age=15724800; includeSubDomains
x-ratelimit-limit-requests:
- '10000'
x-ratelimit-limit-tokens:
- '300000'
x-ratelimit-remaining-requests:
- '9999'
x-ratelimit-remaining-tokens:
- '299761'
x-ratelimit-reset-requests:
- 6ms
x-ratelimit-reset-tokens:
- 47ms
x-request-id:
- 4100fde9c68d27d808de645637b3e7cc
http_version: HTTP/1.1
status_code: 200
version: 1