序言
使用 Prolog 规则生成答案的 LangChain 工具。
概述
PrologTool 类允许生成使用 Prolog 规则生成答案的 langchain 工具。
设置
让我们在文件 family.pl 中使用以下 Prolog 规则:
父级(John、Bianca、Mary)。
家长(John, Bianca, Michael)。
家长(Peter、Patricia、Jennifer)。
partner(X, Y) :- parent(X, Y, _).
#!pip install langchain-prolog
from langchain_prolog import PrologConfig, PrologRunnable, PrologTool
TEST_SCRIPT = "family.pl"
实例
首先创建 Prolog 工具:
schema = PrologRunnable.create_schema("parent", ["men", "women", "child"])
config = PrologConfig(
rules_file=TEST_SCRIPT,
query_schema=schema,
)
prolog_tool = PrologTool(
prolog_config=config,
name="family_query",
description="""
Query family relationships using Prolog.
parent(X, Y, Z) implies only that Z is a child of X and Y.
Input can be a query string like 'parent(john, X, Y)' or 'john, X, Y'"
You have to specify 3 parameters: men, woman, child. Do not use quotes.
""",
)
调用
将 Prolog 工具与 LLM 一起使用并调用函数
#!pip install python-dotenv
from dotenv import find_dotenv, load_dotenv
load_dotenv(find_dotenv(), override=True)
#!pip install langchain-openai
from langchain_core.messages import HumanMessage
from langchain_openai import ChatOpenAI
API 参考:HumanMessage | 聊天OpenAI
要使用该工具,请将其绑定到 LLM 模型:
llm = ChatOpenAI(model="gpt-4o-mini")
llm_with_tools = llm.bind_tools([prolog_tool])
然后查询模型:
query = "Who are John's children?"
messages = [HumanMessage(query)]
response = llm_with_tools.invoke(messages)
LLM 将响应工具调用请求:
messages.append(response)
response.tool_calls[0]
{'name': 'family_query',
'args': {'men': 'john', 'women': None, 'child': None},
'id': 'call_gH8rWamYXITrkfvRP2s5pkbF',
'type': 'tool_call'}
该工具接受此请求并查询 Prolog 数据库:
tool_msg = prolog_tool.invoke(response.tool_calls[0])
该工具将返回一个包含查询的所有解决方案的列表:
messages.append(tool_msg)
tool_msg
ToolMessage(content='[{"Women": "bianca", "Child": "mary"}, {"Women": "bianca", "Child": "michael"}]', name='family_query', tool_call_id='call_gH8rWamYXITrkfvRP2s5pkbF')
然后,我们将其传递给 LLM,LLM 使用工具响应回答原始查询:
answer = llm_with_tools.invoke(messages)
print(answer.content)
John has two children: Mary and Michael, with Bianca as their mother.
链接
将 Prolog Tool 与代理一起使用
要将 prolog 工具与代理一起使用,请将其传递给代理的构造函数:
#!pip install langgraph
from langgraph.prebuilt import create_react_agent
agent_executor = create_react_agent(llm, [prolog_tool])
API 参考:create_react_agent
代理将获取查询并在需要时使用 Prolog 工具:
messages = agent_executor.invoke({"messages": [("human", query)]})
然后,代理收到工具响应并生成答案:
messages["messages"][-1].pretty_print()
==================================[1m Ai Message [0m==================================
John has two children: Mary and Michael, with Bianca as their mother.
API 参考
有关详细信息,请参阅 https://langchain-prolog.readthedocs.io/en/latest/modules.html。