github/langchain
1
0
Fork 0
mirror of https://github.com/hwchase17/langchain.git synced 2025-05-31 12:09:58 +00:00
Code Issues Packages Projects Releases Wiki Activity

[langchain] agents code changes (#15278)

Browse Source 
<!-- Thank you for contributing to LangChain!

Please title your PR "<package>: <description>", where <package> is
whichever of langchain, community, core, experimental, etc. is being
modified.

Replace this entire comment with:
  - **Description:** a description of the change, 
  - **Issue:** the issue # it fixes if applicable,
  - **Dependencies:** any dependencies required for this change,
- **Twitter handle:** we announce bigger features on Twitter. If your PR
gets announced, and you'd like a mention, we'll gladly shout you out!
This commit is contained in:
Harrison Chase 2023-12-28 13:39:08 -08:00 committed by GitHub
parent b86803153e
commit 90aa26a90e
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
14 changed files with 579 additions and 9 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

27
libs/langchain/langchain/agents/__init__.py
View File

@ -56,6 +56,7 @@ from langchain.agents.agent_types import AgentType
from langchain.agents.conversational.base import ConversationalAgent
from langchain.agents.conversational_chat.base import ConversationalChatAgent
from langchain.agents.initialize import initialize_agent
from langchain.agents.json_chat.base import create_json_chat_agent
from langchain.agents.load_tools import (
get_all_tool_names,
load_huggingface_tool,
@ -63,13 +64,24 @@ from langchain.agents.load_tools import (
)
from langchain.agents.loading import load_agent
from langchain.agents.mrkl.base import MRKLChain, ZeroShotAgent
from langchain.agents.openai_functions_agent.base import OpenAIFunctionsAgent
from langchain.agents.openai_functions_agent.base import (
OpenAIFunctionsAgent,
create_openai_functions_agent,
)
from langchain.agents.openai_functions_multi_agent.base import OpenAIMultiFunctionsAgent
from langchain.agents.openai_tools.base import create_openai_tools_agent
from langchain.agents.react.agent import create_react_agent
from langchain.agents.react.base import ReActChain, ReActTextWorldAgent
from langchain.agents.self_ask_with_search.base import SelfAskWithSearchChain
from langchain.agents.structured_chat.base import StructuredChatAgent
from langchain.agents.self_ask_with_search.base import (
SelfAskWithSearchChain,
create_self_ask_with_search_agent,
)
from langchain.agents.structured_chat.base import (
StructuredChatAgent,
create_structured_chat_agent,
)
from langchain.agents.tools import Tool, tool
from langchain.agents.xml.base import XMLAgent
from langchain.agents.xml.base import XMLAgent, create_xml_agent
DEPRECATED_CODE = [
"create_csv_agent",
@ -133,4 +145,11 @@ __all__ = [
"load_tools",
"tool",
"XMLAgent",
"create_openai_functions_agent",
"create_xml_agent",
"create_react_agent",
"create_openai_tools_agent",
"create_self_ask_with_search_agent",
"create_json_chat_agent",
"create_structured_chat_agent",
]

0
libs/langchain/langchain/agents/json_chat/__init__.py Normal file
View File

83
libs/langchain/langchain/agents/json_chat/base.py Normal file
View File

@ -0,0 +1,83 @@
from typing import Sequence
from langchain_core.language_models import BaseLanguageModel
from langchain_core.prompts.chat import ChatPromptTemplate
from langchain_core.runnables import Runnable, RunnablePassthrough
from langchain_core.tools import BaseTool
from langchain.agents.format_scratchpad import format_log_to_messages
from langchain.agents.json_chat.prompt import TEMPLATE_TOOL_RESPONSE
from langchain.agents.output_parsers import JSONAgentOutputParser
from langchain.tools.render import render_text_description
def create_json_chat_agent(
llm: BaseLanguageModel, tools: Sequence[BaseTool], prompt: ChatPromptTemplate
) -> Runnable:
"""Create an agent that uses JSON to format its logic, build for Chat Models.
Examples:
.. code-block:: python
from langchain import hub
from langchain.chat_models import ChatOpenAI
from langchain.agents import AgentExecutor, create_json_chat_agent
prompt = hub.pull("hwchase17/react-chat-json")
model = ChatOpenAI()
tools = ...
agent = create_json_chat_agent(model, tools, prompt)
agent_executor = AgentExecutor(agent=agent, tools=tools)
agent_executor.invoke({"input": "hi"})
# Using with chat history
from langchain_core.messages import AIMessage, HumanMessage
agent_executor.invoke(
{
"input": "what's my name?",
"chat_history": [
HumanMessage(content="hi! my name is bob"),
AIMessage(content="Hello Bob! How can I assist you today?"),
],
}
)
Args:
llm: LLM to use as the agent.
tools: Tools this agent has access to.
prompt: The prompt to use, must have input keys of
`tools`, `tool_names`, and `agent_scratchpad`.
Returns:
A runnable sequence representing an agent. It takes as input all the same input
variables as the prompt passed in does. It returns as output either an
AgentAction or AgentFinish.
"""
missing_vars = {"tools", "tool_names", "agent_scratchpad"}.difference(
prompt.input_variables
)
if missing_vars:
raise ValueError(f"Prompt missing required variables: {missing_vars}")
prompt = prompt.partial(
tools=render_text_description(list(tools)),
tool_names=", ".join([t.name for t in tools]),
)
llm_with_stop = llm.bind(stop=["\nObservation"])
agent = (
RunnablePassthrough.assign(
agent_scratchpad=lambda x: format_log_to_messages(
x["intermediate_steps"], template_tool_response=TEMPLATE_TOOL_RESPONSE
)
)
| prompt
| llm_with_stop
| JSONAgentOutputParser()
)
return agent

9
libs/langchain/langchain/agents/json_chat/prompt.py Normal file
View File

@ -0,0 +1,9 @@
# flake8: noqa
TEMPLATE_TOOL_RESPONSE = """TOOL RESPONSE:
---------------------
{observation}
USER'S INPUT
--------------------
Okay, so what is the response to my last comment? If using information obtained from the tools you must mention it explicitly without mentioning the tool names - I have forgotten all TOOL RESPONSES! Remember to respond with a markdown code snippet of a json blob with a single action, and NOTHING else - even if you just want to respond to the user. Do NOT respond with anything except a JSON snippet no matter what!"""

71
libs/langchain/langchain/agents/openai_functions_agent/base.py
View File

@ -15,6 +15,7 @@ from langchain_core.prompts.chat import (
MessagesPlaceholder,
)
from langchain_core.pydantic_v1 import root_validator
from langchain_core.runnables import Runnable, RunnablePassthrough
from langchain_core.tools import BaseTool
from langchain.agents import BaseSingleActionAgent
@ -226,3 +227,73 @@ class OpenAIFunctionsAgent(BaseSingleActionAgent):
callback_manager=callback_manager,
**kwargs,
)
def create_openai_functions_agent(
llm: BaseLanguageModel, tools: Sequence[BaseTool], prompt: ChatPromptTemplate
) -> Runnable:
"""Create an agent that uses OpenAI function calling.
Examples:
Creating an agent with no memory
.. code-block:: python
from langchain.chat_models import ChatOpenAI
from langchain.agents import AgentExecutor, create_openai_functions_agent
from langchain import hub
prompt = hub.pull("hwchase17/openai-functions-agent")
model = ChatOpenAI()
tools = ...
agent = create_openai_functions_agent(model, tools, prompt)
agent_executor = AgentExecutor(agent=agent, tools=tools)
agent_executor.invoke({"input": "hi"})
# Using with chat history
from langchain_core.messages import AIMessage, HumanMessage
agent_executor.invoke(
{
"input": "what's my name?",
"chat_history": [
HumanMessage(content="hi! my name is bob"),
AIMessage(content="Hello Bob! How can I assist you today?"),
],
}
)
Args:
llm: LLM to use as the agent. Should work with OpenAI function calling,
so either be an OpenAI model that supports that or a wrapper of
a different model that adds in equivalent support.
tools: Tools this agent has access to.
prompt: The prompt to use, must have an input key of `agent_scratchpad`.
Returns:
A runnable sequence representing an agent. It takes as input all the same input
variables as the prompt passed in does. It returns as output either an
AgentAction or AgentFinish.
"""
if "agent_scratchpad" not in prompt.input_variables:
raise ValueError(
"Prompt must have input variable `agent_scratchpad`, but wasn't found. "
f"Found {prompt.input_variables} instead."
)
llm_with_tools = llm.bind(
functions=[format_tool_to_openai_function(t) for t in tools]
)
agent = (
RunnablePassthrough.assign(
agent_scratchpad=lambda x: format_to_openai_function_messages(
x["intermediate_steps"]
)
)
| prompt
| llm_with_tools
| OpenAIFunctionsAgentOutputParser()
)
return agent

0
libs/langchain/langchain/agents/openai_tools/__init__.py Normal file
View File

79
libs/langchain/langchain/agents/openai_tools/base.py Normal file
View File

@ -0,0 +1,79 @@
from typing import Sequence
from langchain_core.language_models import BaseLanguageModel
from langchain_core.prompts.chat import ChatPromptTemplate
from langchain_core.runnables import Runnable, RunnablePassthrough
from langchain_core.tools import BaseTool
from langchain.agents.format_scratchpad.openai_tools import (
format_to_openai_tool_messages,
)
from langchain.agents.output_parsers.openai_tools import OpenAIToolsAgentOutputParser
from langchain.tools.render import format_tool_to_openai_tool
def create_openai_tools_agent(
llm: BaseLanguageModel, tools: Sequence[BaseTool], prompt: ChatPromptTemplate
) -> Runnable:
"""Create an agent that uses OpenAI tools.
Examples:
.. code-block:: python
from langchain import hub
from langchain.chat_models import ChatOpenAI
from langchain.agents import AgentExecutor, create_openai_tools_agent
prompt = hub.pull("hwchase17/openai-tools-agent")
model = ChatOpenAI()
tools = ...
agent = create_openai_tools_agent(model, tools, prompt)
agent_executor = AgentExecutor(agent=agent, tools=tools)
agent_executor.invoke({"input": "hi"})
# Using with chat history
from langchain_core.messages import AIMessage, HumanMessage
agent_executor.invoke(
{
"input": "what's my name?",
"chat_history": [
HumanMessage(content="hi! my name is bob"),
AIMessage(content="Hello Bob! How can I assist you today?"),
],
}
)
Args:
llm: LLM to use as the agent.
tools: Tools this agent has access to.
prompt: The prompt to use, must have input keys of `agent_scratchpad`.
Returns:
A runnable sequence representing an agent. It takes as input all the same input
variables as the prompt passed in does. It returns as output either an
AgentAction or AgentFinish.
"""
missing_vars = {"agent_scratchpad"}.difference(prompt.input_variables)
if missing_vars:
raise ValueError(f"Prompt missing required variables: {missing_vars}")
llm_with_tools = llm.bind(
tools=[format_tool_to_openai_tool(tool) for tool in tools]
)
agent = (
RunnablePassthrough.assign(
agent_scratchpad=lambda x: format_to_openai_tool_messages(
x["intermediate_steps"]
)
)
| prompt
| llm_with_tools
| OpenAIToolsAgentOutputParser()
)
return agent

79
libs/langchain/langchain/agents/react/agent.py Normal file
View File

@ -0,0 +1,79 @@
from __future__ import annotations
from typing import Sequence
from langchain_core.language_models import BaseLanguageModel
from langchain_core.prompts import BasePromptTemplate
from langchain_core.runnables import Runnable, RunnablePassthrough
from langchain_core.tools import BaseTool
from langchain.agents.format_scratchpad import format_log_to_str
from langchain.agents.output_parsers import ReActSingleInputOutputParser
from langchain.tools.render import render_text_description
def create_react_agent(
llm: BaseLanguageModel, tools: Sequence[BaseTool], prompt: BasePromptTemplate
) -> Runnable:
"""Create an agent that uses ReAct prompting.
Examples:
.. code-block:: python
from langchain import hub
from langchain.llms import OpenAI
from langchain.agents import AgentExecutor, create_react_agent
prompt = hub.pull("hwchase17/react")
model = OpenAI()
tools = ...
agent = create_react_agent(model, tools, prompt)
agent_executor = AgentExecutor(agent=agent, tools=tools)
agent_executor.invoke({"input": "hi"})
# Use with chat history
from langchain_core.messages import AIMessage, HumanMessage
agent_executor.invoke(
{
"input": "what's my name?",
# Notice that chat_history is a string
# since this prompt is aimed at LLMs, not chat models
"chat_history": "Human: My name is Bob\nAI: Hello Bob!",
}
)
Args:
llm: LLM to use as the agent.
tools: Tools this agent has access to.
prompt: The prompt to use, must have input keys of
`tools`, `tool_names`, and `agent_scratchpad`.
Returns:
A runnable sequence representing an agent. It takes as input all the same input
variables as the prompt passed in does. It returns as output either an
AgentAction or AgentFinish.
"""
missing_vars = {"tools", "tool_names", "agent_scratchpad"}.difference(
prompt.input_variables
)
if missing_vars:
raise ValueError(f"Prompt missing required variables: {missing_vars}")
prompt = prompt.partial(
tools=render_text_description(list(tools)),
tool_names=", ".join([t.name for t in tools]),
)
llm_with_stop = llm.bind(stop=["\nObservation"])
agent = (
RunnablePassthrough.assign(
agent_scratchpad=lambda x: format_log_to_str(x["intermediate_steps"]),
)
| prompt
| llm_with_stop
| ReActSingleInputOutputParser()
)
return agent

69
libs/langchain/langchain/agents/self_ask_with_search/base.py
View File

@ -4,10 +4,12 @@ from typing import Any, Sequence, Union
from langchain_core.language_models import BaseLanguageModel
from langchain_core.prompts import BasePromptTemplate
from langchain_core.pydantic_v1 import Field
from langchain_core.runnables import Runnable, RunnablePassthrough
from langchain_core.tools import BaseTool
from langchain.agents.agent import Agent, AgentExecutor, AgentOutputParser
from langchain.agents.agent_types import AgentType
from langchain.agents.format_scratchpad import format_log_to_str
from langchain.agents.self_ask_with_search.output_parser import SelfAskOutputParser
from langchain.agents.self_ask_with_search.prompt import PROMPT
from langchain.agents.tools import Tool
@ -79,3 +81,70 @@ class SelfAskWithSearchChain(AgentExecutor):
)
agent = SelfAskWithSearchAgent.from_llm_and_tools(llm, [search_tool])
super().__init__(agent=agent, tools=[search_tool], **kwargs)
def create_self_ask_with_search_agent(
llm: BaseLanguageModel, tools: Sequence[BaseTool], prompt: BasePromptTemplate
) -> Runnable:
"""Create an agent that uses self-ask with search prompting.
Examples:
.. code-block:: python
from langchain import hub
from langchain.chat_models import ChatAnthropic
from langchain.agents import (
AgentExecutor, create_self_ask_with_search_agent
)
prompt = hub.pull("hwchase17/self-ask-with-search")
model = ChatAnthropic()
tools = [...] # Should just be one tool with name `Intermediate Answer`
agent = create_self_ask_with_search_agent(model, tools, prompt)
agent_executor = AgentExecutor(agent=agent, tools=tools)
agent_executor.invoke({"input": "hi"})
Args:
llm: LLM to use as the agent.
tools: List of tools. Should just be of length 1, with that tool having
name `Intermediate Answer`
prompt: The prompt to use, must have input keys of `agent_scratchpad`.
Returns:
A runnable sequence representing an agent. It takes as input all the same input
variables as the prompt passed in does. It returns as output either an
AgentAction or AgentFinish.
"""
missing_vars = {"agent_scratchpad"}.difference(prompt.input_variables)
if missing_vars:
raise ValueError(f"Prompt missing required variables: {missing_vars}")
if len(tools) != 1:
raise ValueError("This agent expects exactly one tool")
tool = list(tools)[0]
if tool.name != "Intermediate Answer":
raise ValueError(
"This agent expects the tool to be named `Intermediate Answer`"
)
llm_with_stop = llm.bind(stop=["\nIntermediate answer:"])
agent = (
RunnablePassthrough.assign(
agent_scratchpad=lambda x: format_log_to_str(
x["intermediate_steps"],
observation_prefix="\nIntermediate answer: ",
llm_prefix="",
),
# Give it a default
chat_history=lambda x: x.get("chat_history", ""),
)
| prompt
| llm_with_stop
| SelfAskOutputParser()
)
return agent

74
libs/langchain/langchain/agents/structured_chat/base.py
View File

@ -10,8 +10,11 @@ from langchain_core.prompts.chat import (
SystemMessagePromptTemplate,
)
from langchain_core.pydantic_v1 import Field
from langchain_core.runnables import Runnable, RunnablePassthrough
from langchain.agents.agent import Agent, AgentOutputParser
from langchain.agents.format_scratchpad import format_log_to_str
from langchain.agents.output_parsers import JSONAgentOutputParser
from langchain.agents.structured_chat.output_parser import (
StructuredChatOutputParserWithRetries,
)
@ -19,6 +22,7 @@ from langchain.agents.structured_chat.prompt import FORMAT_INSTRUCTIONS, PREFIX,
from langchain.callbacks.base import BaseCallbackManager
from langchain.chains.llm import LLMChain
from langchain.tools import BaseTool
from langchain.tools.render import render_text_description_and_args
HUMAN_MESSAGE_TEMPLATE = "{input}\n\n{agent_scratchpad}"
@ -142,3 +146,73 @@ class StructuredChatAgent(Agent):
@property
def _agent_type(self) -> str:
raise ValueError
def create_structured_chat_agent(
llm: BaseLanguageModel, tools: Sequence[BaseTool], prompt: ChatPromptTemplate
) -> Runnable:
"""Create an agent aimed at supporting tools with multiple inputs.
Examples:
.. code-block:: python
from langchain import hub
from langchain.chat_models import ChatOpenAI
from langchain.agents import AgentExecutor, create_structured_chat_agent
prompt = hub.pull("hwchase17/structured-chat-agent")
model = ChatOpenAI()
tools = ...
agent = create_structured_chat_agent(model, tools, prompt)
agent_executor = AgentExecutor(agent=agent, tools=tools)
agent_executor.invoke({"input": "hi"})
# Using with chat history
from langchain_core.messages import AIMessage, HumanMessage
agent_executor.invoke(
{
"input": "what's my name?",
"chat_history": [
HumanMessage(content="hi! my name is bob"),
AIMessage(content="Hello Bob! How can I assist you today?"),
],
}
)
Args:
llm: LLM to use as the agent.
tools: Tools this agent has access to.
prompt: The prompt to use, must have input keys of
`tools`, `tool_names`, and `agent_scratchpad`.
Returns:
A runnable sequence representing an agent. It takes as input all the same input
variables as the prompt passed in does. It returns as output either an
AgentAction or AgentFinish.
"""
missing_vars = {"tools", "tool_names", "agent_scratchpad"}.difference(
prompt.input_variables
)
if missing_vars:
raise ValueError(f"Prompt missing required variables: {missing_vars}")
prompt = prompt.partial(
tools=render_text_description_and_args(list(tools)),
tool_names=", ".join([t.name for t in tools]),
)
llm_with_stop = llm.bind(stop=["Observation"])
agent = (
RunnablePassthrough.assign(
agent_scratchpad=lambda x: format_log_to_str(x["intermediate_steps"]),
)
| prompt
| llm_with_stop
| JSONAgentOutputParser()
)
return agent

82
libs/langchain/langchain/agents/xml/base.py
View File

@ -1,14 +1,19 @@
from typing import Any, List, Tuple, Union
from typing import Any, List, Sequence, Tuple, Union
from langchain_core.agents import AgentAction, AgentFinish
from langchain_core.language_models import BaseLanguageModel
from langchain_core.prompts.base import BasePromptTemplate
from langchain_core.prompts.chat import AIMessagePromptTemplate, ChatPromptTemplate
from langchain_core.runnables import Runnable, RunnablePassthrough
from langchain_core.tools import BaseTool
from langchain.agents.agent import BaseSingleActionAgent
from langchain.agents.output_parsers.xml import XMLAgentOutputParser
from langchain.agents.format_scratchpad import format_xml
from langchain.agents.output_parsers import XMLAgentOutputParser
from langchain.agents.xml.prompt import agent_instructions
from langchain.callbacks.base import Callbacks
from langchain.chains.llm import LLMChain
from langchain.tools.render import render_text_description
class XMLAgent(BaseSingleActionAgent):
@ -42,9 +47,10 @@ class XMLAgent(BaseSingleActionAgent):
@staticmethod
def get_default_prompt() -> ChatPromptTemplate:
return ChatPromptTemplate.from_template(
agent_instructions
) + AIMessagePromptTemplate.from_template("{intermediate_steps}")
base_prompt = ChatPromptTemplate.from_template(agent_instructions)
return base_prompt + AIMessagePromptTemplate.from_template(
"{intermediate_steps}"
)
@staticmethod
def get_default_output_parser() -> XMLAgentOutputParser:
@ -97,3 +103,69 @@ class XMLAgent(BaseSingleActionAgent):
}
response = await self.llm_chain.acall(inputs, callbacks=callbacks)
return response[self.llm_chain.output_key]
def create_xml_agent(
llm: BaseLanguageModel, tools: Sequence[BaseTool], prompt: BasePromptTemplate
) -> Runnable:
"""Create an agent that uses XML to format its logic.
Examples:
.. code-block:: python
from langchain import hub
from langchain.chat_models import ChatAnthropic
from langchain.agents import AgentExecutor, create_xml_agent
prompt = hub.pull("hwchase17/xml-agent-convo")
model = ChatAnthropic()
tools = ...
agent = create_xml_agent(model, tools, prompt)
agent_executor = AgentExecutor(agent=agent, tools=tools)
agent_executor.invoke({"input": "hi"})
# Use with chat history
from langchain_core.messages import AIMessage, HumanMessage
agent_executor.invoke(
{
"input": "what's my name?",
# Notice that chat_history is a string
# since this prompt is aimed at LLMs, not chat models
"chat_history": "Human: My name is Bob\nAI: Hello Bob!",
}
)
Args:
llm: LLM to use as the agent.
tools: Tools this agent has access to.
prompt: The prompt to use, must have input keys of
`tools` and `agent_scratchpad`.
Returns:
A runnable sequence representing an agent. It takes as input all the same input
variables as the prompt passed in does. It returns as output either an
AgentAction or AgentFinish.
"""
missing_vars = {"tools", "agent_scratchpad"}.difference(prompt.input_variables)
if missing_vars:
raise ValueError(f"Prompt missing required variables: {missing_vars}")
prompt = prompt.partial(
tools=render_text_description(list(tools)),
)
llm_with_stop = llm.bind(stop=["</tool_input>"])
agent = (
RunnablePassthrough.assign(
agent_scratchpad=lambda x: format_xml(x["intermediate_steps"]),
)
| prompt
| llm_with_stop
| XMLAgentOutputParser()
)
return agent

1
libs/langchain/langchain/agents/xml/prompt.py
View File

@ -1,4 +1,5 @@
# flake8: noqa
# TODO: deprecate
agent_instructions = """You are a helpful assistant. Help the user answer any questions.
You have access to the following tools:

7
libs/langchain/tests/unit_tests/agents/test_imports.py
View File

@ -35,6 +35,13 @@ EXPECTED_ALL = [
"load_tools",
"tool",
"XMLAgent",
"create_openai_functions_agent",
"create_xml_agent",
"create_react_agent",
"create_openai_tools_agent",
"create_self_ask_with_search_agent",
"create_json_chat_agent",
"create_structured_chat_agent",
]

7
libs/langchain/tests/unit_tests/agents/test_public_api.py
View File

@ -35,6 +35,13 @@ _EXPECTED = [
"load_huggingface_tool",
"load_tools",
"tool",
"create_openai_functions_agent",
"create_xml_agent",
"create_react_agent",
"create_openai_tools_agent",
"create_self_ask_with_search_agent",
"create_json_chat_agent",
"create_structured_chat_agent",
]