构建基于SQL数据的问答系统

先决条件

本指南假定您熟悉以下概念

• 聊天模型
• 工具
• 代理
• LangGraph

使LLM系统能够查询结构化数据在性质上可能与非结构化文本数据有所不同。后者通常生成可针对向量数据库进行搜索的文本，而结构化数据的方法通常是让LLM编写并执行DSL（如SQL）查询。在本指南中，我们将介绍在数据库中创建表格数据问答系统的基本方法。我们将介绍使用链和代理的实现。这些系统将允许我们提出关于数据库数据的问题，并获得自然语言答案。两者之间的主要区别在于，我们的代理可以根据需要多次循环查询数据库以回答问题。

⚠️ 安全注意事项 ⚠️

构建SQL数据库的问答系统需要执行模型生成的SQL查询。这样做存在固有风险。请确保您的数据库连接权限始终尽可能地窄，以满足链/代理的需求。这将减轻但不能消除构建模型驱动系统的风险。有关一般安全最佳实践的更多信息，请参见此处。

架构

从高层次来看，这些系统的步骤是：

1. 将问题转换为SQL查询：模型将用户输入转换为SQL查询。
2. 执行SQL查询：执行查询。
3. 回答问题：模型使用查询结果响应用户输入。

请注意，查询CSV数据可以遵循类似的方法。有关CSV数据问答的更多详细信息，请参阅我们的操作指南。

设置

首先，获取所需的包并设置环境变量

%%capture --no-stderr
%pip install --upgrade --quiet langchain-community langgraph

# Comment out the below to opt-out of using LangSmith in this notebook. Not required.
if not os.environ.get("LANGSMITH_API_KEY"):
    os.environ["LANGSMITH_API_KEY"] = getpass.getpass()
    os.environ["LANGSMITH_TRACING"] = "true"

示例数据

下面的示例将使用SQLite连接到Chinook数据库，这是一个代表数字媒体商店的示例数据库。按照这些安装步骤，在与此notebook相同的目录中创建Chinook.db。您也可以通过命令行下载并构建数据库

curl -s https://raw.githubusercontent.com/lerocha/chinook-database/master/ChinookDatabase/DataSources/Chinook_Sqlite.sql | sqlite3 Chinook.db

现在，Chinook.db已在我们的目录中，我们可以使用 SQLAlchemy 驱动的 SQLDatabase 类与其交互。

from langchain_community.utilities import SQLDatabase

db = SQLDatabase.from_uri("sqlite:///Chinook.db")
print(db.dialect)
print(db.get_usable_table_names())
db.run("SELECT * FROM Artist LIMIT 10;")

API 参考：SQLDatabase

sqlite
['Album', 'Artist', 'Customer', 'Employee', 'Genre', 'Invoice', 'InvoiceLine', 'MediaType', 'Playlist', 'PlaylistTrack', 'Track']

"[(1, 'AC/DC'), (2, 'Accept'), (3, 'Aerosmith'), (4, 'Alanis Morissette'), (5, 'Alice In Chains'), (6, 'Antônio Carlos Jobim'), (7, 'Apocalyptica'), (8, 'Audioslave'), (9, 'BackBeat'), (10, 'Billy Cobham')]"

太好了！我们有一个可以查询的SQL数据库。现在让我们尝试将其连接到LLM。

链

链是可预测步骤的组合。在LangGraph中，我们可以通过简单的节点序列来表示一个链。让我们创建一系列步骤，给定一个问题，执行以下操作：

• 将问题转换为SQL查询；
• 执行查询；
• 使用结果回答原始问题。

这种安排不支持某些场景。例如，该系统将对任何用户输入（甚至是“你好”）执行SQL查询。重要的是，正如我们将在下面看到的，有些问题需要多次查询才能回答。我们将在“代理”部分讨论这些场景。

应用程序状态

LangGraph应用程序的状态控制着哪些数据作为输入进入应用程序、在步骤之间传输以及由应用程序输出。它通常是TypedDict，但也可以是Pydantic BaseModel。

对于此应用程序，我们只需跟踪输入问题、生成的查询、查询结果和生成的答案。

from typing_extensions import TypedDict


class State(TypedDict):
    question: str
    query: str
    result: str
    answer: str

现在我们只需要能够操作此状态并填充其内容的函数。

将问题转换为SQL查询

第一步是获取用户输入并将其转换为SQL查询。为了可靠地获取SQL查询（没有markdown格式和解释或澄清），我们将利用LangChain的结构化输出抽象。

让我们为应用程序选择一个聊天模型

选择 聊天模型

OpenAI

Anthropic

Azure

Google Gemini

Google Vertex

AWS

Groq

Cohere

NVIDIA

Fireworks AI

Mistral AI

Together AI

IBM watsonx

Databricks

xAI

Perplexity

pip install -qU "langchain[google-genai]"

import getpass
import os

if not os.environ.get("GOOGLE_API_KEY"):
  os.environ["GOOGLE_API_KEY"] = getpass.getpass("Enter API key for Google Gemini: ")

from langchain.chat_models import init_chat_model

llm = init_chat_model("gemini-2.0-flash", model_provider="google_genai")

让我们为模型提供一些指令

from langchain_core.prompts import ChatPromptTemplate

system_message = """
Given an input question, create a syntactically correct {dialect} query to
run to help find the answer. Unless the user specifies in his question a
specific number of examples they wish to obtain, always limit your query to
at most {top_k} results. You can order the results by a relevant column to
return the most interesting examples in the database.

Never query for all the columns from a specific table, only ask for a the
few relevant columns given the question.

Pay attention to use only the column names that you can see in the schema
description. Be careful to not query for columns that do not exist. Also,
pay attention to which column is in which table.

Only use the following tables:
{table_info}
"""

user_prompt = "Question: {input}"

query_prompt_template = ChatPromptTemplate(
    [("system", system_message), ("user", user_prompt)]
)

for message in query_prompt_template.messages:
    message.pretty_print()

API 参考：ChatPromptTemplate

================================ System Message ================================


Given an input question, create a syntactically correct {dialect} query to
run to help find the answer. Unless the user specifies in his question a
specific number of examples they wish to obtain, always limit your query to
at most {top_k} results. You can order the results by a relevant column to
return the most interesting examples in the database.

Never query for all the columns from a specific table, only ask for a the
few relevant columns given the question.

Pay attention to use only the column names that you can see in the schema
description. Be careful to not query for columns that do not exist. Also,
pay attention to which column is in which table.

Only use the following tables:
{table_info}

================================ Human Message =================================

Question: {input}

提示包括我们需要填充的几个参数，例如SQL方言和表模式。LangChain的SQLDatabase对象包含有助于此的方法。我们的write_query步骤将填充这些参数并提示模型生成SQL查询。

from typing_extensions import Annotated


class QueryOutput(TypedDict):
    """Generated SQL query."""

    query: Annotated[str, ..., "Syntactically valid SQL query."]


def write_query(state: State):
    """Generate SQL query to fetch information."""
    prompt = query_prompt_template.invoke(
        {
            "dialect": db.dialect,
            "top_k": 10,
            "table_info": db.get_table_info(),
            "input": state["question"],
        }
    )
    structured_llm = llm.with_structured_output(QueryOutput)
    result = structured_llm.invoke(prompt)
    return {"query": result["query"]}

让我们测试一下

write_query({"question": "How many Employees are there?"})

{'query': 'SELECT COUNT(*) as employee_count FROM Employee;'}

执行查询

这是创建SQL链最危险的部分。 请仔细考虑是否可以在您的数据上运行自动化查询。尽可能最小化数据库连接权限。在查询执行之前，考虑在您的链中添加人工审批步骤（见下文）。

为了执行查询，我们将从langchain-community加载一个工具。我们的execute_query节点将只封装此工具。

from langchain_community.tools.sql_database.tool import QuerySQLDatabaseTool


def execute_query(state: State):
    """Execute SQL query."""
    execute_query_tool = QuerySQLDatabaseTool(db=db)
    return {"result": execute_query_tool.invoke(state["query"])}

API 参考：QuerySQLDatabaseTool

测试此步骤

execute_query({"query": "SELECT COUNT(EmployeeId) AS EmployeeCount FROM Employee;"})

{'result': '[(8,)]'}

生成答案

最后，我们的最后一步会根据从数据库中提取的信息生成问题的答案。

def generate_answer(state: State):
    """Answer question using retrieved information as context."""
    prompt = (
        "Given the following user question, corresponding SQL query, "
        "and SQL result, answer the user question.\n\n"
        f"Question: {state['question']}\n"
        f"SQL Query: {state['query']}\n"
        f"SQL Result: {state['result']}"
    )
    response = llm.invoke(prompt)
    return {"answer": response.content}

使用LangGraph进行编排

最后，我们将应用程序编译成一个单独的graph对象。在这种情况下，我们只是将这三个步骤连接成一个单一的序列。

from langgraph.graph import START, StateGraph

graph_builder = StateGraph(State).add_sequence(
    [write_query, execute_query, generate_answer]
)
graph_builder.add_edge(START, "write_query")
graph = graph_builder.compile()

API 参考：StateGraph

LangGraph 还内置了用于可视化应用程序控制流的实用程序。

from IPython.display import Image, display

display(Image(graph.get_graph().draw_mermaid_png()))

让我们测试一下我们的应用程序！请注意，我们可以流式传输单个步骤的结果。

for step in graph.stream(
    {"question": "How many employees are there?"}, stream_mode="updates"
):
    print(step)

{'write_query': {'query': 'SELECT COUNT(*) as employee_count FROM Employee;'}}
{'execute_query': {'result': '[(8,)]'}}
{'generate_answer': {'answer': 'There are 8 employees in total.'}}

查看LangSmith 追踪。

人工干预

LangGraph支持许多对此工作流有用的功能。其中之一是人工干预：我们可以在敏感步骤（例如SQL查询的执行）之前中断应用程序以进行人工审查。这得益于LangGraph的持久化层，它将运行进度保存到您选择的存储中。下面，我们指定内存存储。

from langgraph.checkpoint.memory import MemorySaver

memory = MemorySaver()
graph = graph_builder.compile(checkpointer=memory, interrupt_before=["execute_query"])

# Now that we're using persistence, we need to specify a thread ID
# so that we can continue the run after review.
config = {"configurable": {"thread_id": "1"}}

API 参考：MemorySaver

display(Image(graph.get_graph().draw_mermaid_png()))

让我们重复相同的运行，添加一个简单的“是/否”审批步骤。

for step in graph.stream(
    {"question": "How many employees are there?"},
    config,
    stream_mode="updates",
):
    print(step)

try:
    user_approval = input("Do you want to go to execute query? (yes/no): ")
except Exception:
    user_approval = "no"

if user_approval.lower() == "yes":
    # If approved, continue the graph execution
    for step in graph.stream(None, config, stream_mode="updates"):
        print(step)
else:
    print("Operation cancelled by user.")

{'write_query': {'query': 'SELECT COUNT(EmployeeId) AS EmployeeCount FROM Employee;'}}
{'__interrupt__': ()}
``````output
Do you want to go to execute query? (yes/no):  yes
``````output
{'execute_query': {'result': '[(8,)]'}}
{'generate_answer': {'answer': 'There are 8 employees.'}}

有关更多详细信息和示例，请参阅此LangGraph指南。

下一步

对于更复杂的查询生成，我们可能需要创建少样本提示或添加查询检查步骤。有关此类及更多高级技术，请查看：

• 提示策略：高级提示工程技术。
• 查询检查：添加查询验证和错误处理。
• 大型数据库：处理大型数据库的技术。

代理

代理利用LLM的推理能力在执行期间做出决策。使用代理可以将查询生成和执行过程中的额外判断权交给它们。尽管它们的行为不如上述“链”可预测，但它们具有一些优势：

• 它们可以根据需要多次查询数据库以回答用户问题。
• 它们可以通过运行生成的查询、捕获回溯并正确地重新生成来从错误中恢复。
• 它们可以根据数据库的模式以及数据库的内容（例如描述特定表）回答问题。

下面我们组装一个最小化的SQL代理。我们将使用LangChain的SQLDatabaseToolkit为其配备一套工具。使用LangGraph的预构建ReAct代理构造器，我们可以在一行代码中完成此操作。

提示

请参阅LangGraph的SQL代理教程，了解更高级的SQL代理公式。

SQLDatabaseToolkit包含可以执行以下操作的工具：

• 创建并执行查询
• 检查查询语法
• 检索表描述
• ......等等

from langchain_community.agent_toolkits import SQLDatabaseToolkit

toolkit = SQLDatabaseToolkit(db=db, llm=llm)

tools = toolkit.get_tools()

tools

API 参考：SQLDatabaseToolkit

[QuerySQLDatabaseTool(description="Input to this tool is a detailed and correct SQL query, output is a result from the database. If the query is not correct, an error message will be returned. If an error is returned, rewrite the query, check the query, and try again. If you encounter an issue with Unknown column 'xxxx' in 'field list', use sql_db_schema to query the correct table fields.", db=<langchain_community.utilities.sql_database.SQLDatabase object at 0x10d5f9120>),
 InfoSQLDatabaseTool(description='Input to this tool is a comma-separated list of tables, output is the schema and sample rows for those tables. Be sure that the tables actually exist by calling sql_db_list_tables first! Example Input: table1, table2, table3', db=<langchain_community.utilities.sql_database.SQLDatabase object at 0x10d5f9120>),
 ListSQLDatabaseTool(db=<langchain_community.utilities.sql_database.SQLDatabase object at 0x10d5f9120>),
 QuerySQLCheckerTool(description='Use this tool to double check if your query is correct before executing it. Always use this tool before executing a query with sql_db_query!', db=<langchain_community.utilities.sql_database.SQLDatabase object at 0x10d5f9120>, llm=ChatOpenAI(client=<openai.resources.chat.completions.Completions object at 0x119315480>, async_client=<openai.resources.chat.completions.AsyncCompletions object at 0x119317550>, root_client=<openai.OpenAI object at 0x10d5f8df0>, root_async_client=<openai.AsyncOpenAI object at 0x1193154e0>, model_name='gpt-4o', temperature=0.0, model_kwargs={}, openai_api_key=SecretStr('**********')), llm_chain=LLMChain(verbose=False, prompt=PromptTemplate(input_variables=['dialect', 'query'], input_types={}, partial_variables={}, template='\n{query}\nDouble check the {dialect} query above for common mistakes, including:\n- Using NOT IN with NULL values\n- Using UNION when UNION ALL should have been used\n- Using BETWEEN for exclusive ranges\n- Data type mismatch in predicates\n- Properly quoting identifiers\n- Using the correct number of arguments for functions\n- Casting to the correct data type\n- Using the proper columns for joins\n\nIf there are any of the above mistakes, rewrite the query. If there are no mistakes, just reproduce the original query.\n\nOutput the final SQL query only.\n\nSQL Query: '), llm=ChatOpenAI(client=<openai.resources.chat.completions.Completions object at 0x119315480>, async_client=<openai.resources.chat.completions.AsyncCompletions object at 0x119317550>, root_client=<openai.OpenAI object at 0x10d5f8df0>, root_async_client=<openai.AsyncOpenAI object at 0x1193154e0>, model_name='gpt-4o', temperature=0.0, model_kwargs={}, openai_api_key=SecretStr('**********')), output_parser=StrOutputParser(), llm_kwargs={}))]

系统提示

我们还需要为代理加载一个系统提示。这将包含关于如何行为的指令。请注意，下面的提示有几个参数，我们将在下面进行赋值。

system_message = """
You are an agent designed to interact with a SQL database.
Given an input question, create a syntactically correct {dialect} query to run,
then look at the results of the query and return the answer. Unless the user
specifies a specific number of examples they wish to obtain, always limit your
query to at most {top_k} results.

You can order the results by a relevant column to return the most interesting
examples in the database. Never query for all the columns from a specific table,
only ask for the relevant columns given the question.

You MUST double check your query before executing it. If you get an error while
executing a query, rewrite the query and try again.

DO NOT make any DML statements (INSERT, UPDATE, DELETE, DROP etc.) to the
database.

To start you should ALWAYS look at the tables in the database to see what you
can query. Do NOT skip this step.

Then you should query the schema of the most relevant tables.
""".format(
    dialect="SQLite",
    top_k=5,
)

初始化代理

我们将使用预构建的LangGraph代理来构建我们的代理。

from langchain_core.messages import HumanMessage
from langgraph.prebuilt import create_react_agent

agent_executor = create_react_agent(llm, tools, prompt=system_message)

API 参考：HumanMessage | create_react_agent

考虑代理如何响应以下问题

question = "Which country's customers spent the most?"

for step in agent_executor.stream(
    {"messages": [{"role": "user", "content": question}]},
    stream_mode="values",
):
    step["messages"][-1].pretty_print()

================================ Human Message =================================

Which country's customers spent the most?
================================== Ai Message ==================================
Tool Calls:
  sql_db_list_tables (call_tFp7HYD6sAAmCShgeqkVZH6Q)
 Call ID: call_tFp7HYD6sAAmCShgeqkVZH6Q
  Args:
================================= Tool Message =================================
Name: sql_db_list_tables

Album, Artist, Customer, Employee, Genre, Invoice, InvoiceLine, MediaType, Playlist, PlaylistTrack, Track
================================== Ai Message ==================================
Tool Calls:
  sql_db_schema (call_KJZ1Jx6JazyDdJa0uH1UeiOz)
 Call ID: call_KJZ1Jx6JazyDdJa0uH1UeiOz
  Args:
    table_names: Customer, Invoice
================================= Tool Message =================================
Name: sql_db_schema


CREATE TABLE "Customer" (
	"CustomerId" INTEGER NOT NULL, 
	"FirstName" NVARCHAR(40) NOT NULL, 
	"LastName" NVARCHAR(20) NOT NULL, 
	"Company" NVARCHAR(80), 
	"Address" NVARCHAR(70), 
	"City" NVARCHAR(40), 
	"State" NVARCHAR(40), 
	"Country" NVARCHAR(40), 
	"PostalCode" NVARCHAR(10), 
	"Phone" NVARCHAR(24), 
	"Fax" NVARCHAR(24), 
	"Email" NVARCHAR(60) NOT NULL, 
	"SupportRepId" INTEGER, 
	PRIMARY KEY ("CustomerId"), 
	FOREIGN KEY("SupportRepId") REFERENCES "Employee" ("EmployeeId")
)

/*
3 rows from Customer table:
CustomerId	FirstName	LastName	Company	Address	City	State	Country	PostalCode	Phone	Fax	Email	SupportRepId
1	Luís	Gonçalves	Embraer - Empresa Brasileira de Aeronáutica S.A.	Av. Brigadeiro Faria Lima, 2170	São José dos Campos	SP	Brazil	12227-000	+55 (12) 3923-5555	+55 (12) 3923-5566	luisg@embraer.com.br	3
2	Leonie	Köhler	None	Theodor-Heuss-Straße 34	Stuttgart	None	Germany	70174	+49 0711 2842222	None	leonekohler@surfeu.de	5
3	François	Tremblay	None	1498 rue Bélanger	Montréal	QC	Canada	H2G 1A7	+1 (514) 721-4711	None	ftremblay@gmail.com	3
*/


CREATE TABLE "Invoice" (
	"InvoiceId" INTEGER NOT NULL, 
	"CustomerId" INTEGER NOT NULL, 
	"InvoiceDate" DATETIME NOT NULL, 
	"BillingAddress" NVARCHAR(70), 
	"BillingCity" NVARCHAR(40), 
	"BillingState" NVARCHAR(40), 
	"BillingCountry" NVARCHAR(40), 
	"BillingPostalCode" NVARCHAR(10), 
	"Total" NUMERIC(10, 2) NOT NULL, 
	PRIMARY KEY ("InvoiceId"), 
	FOREIGN KEY("CustomerId") REFERENCES "Customer" ("CustomerId")
)

/*
3 rows from Invoice table:
InvoiceId	CustomerId	InvoiceDate	BillingAddress	BillingCity	BillingState	BillingCountry	BillingPostalCode	Total
1	2	2021-01-01 00:00:00	Theodor-Heuss-Straße 34	Stuttgart	None	Germany	70174	1.98
2	4	2021-01-02 00:00:00	Ullevålsveien 14	Oslo	None	Norway	0171	3.96
3	8	2021-01-03 00:00:00	Grétrystraat 63	Brussels	None	Belgium	1000	5.94
*/
================================== Ai Message ==================================
Tool Calls:
  sql_db_query_checker (call_AQuTGbgH63u4gPgyV723yrjX)
 Call ID: call_AQuTGbgH63u4gPgyV723yrjX
  Args:
    query: SELECT c.Country, SUM(i.Total) as TotalSpent FROM Customer c JOIN Invoice i ON c.CustomerId = i.CustomerId GROUP BY c.Country ORDER BY TotalSpent DESC LIMIT 1;
================================= Tool Message =================================
Name: sql_db_query_checker

\`\`\`sql
SELECT c.Country, SUM(i.Total) as TotalSpent FROM Customer c JOIN Invoice i ON c.CustomerId = i.CustomerId GROUP BY c.Country ORDER BY TotalSpent DESC LIMIT 1;
\`\`\`
================================== Ai Message ==================================
Tool Calls:
  sql_db_query (call_B88EwU44nwwpQL5M9nlcemSU)
 Call ID: call_B88EwU44nwwpQL5M9nlcemSU
  Args:
    query: SELECT c.Country, SUM(i.Total) as TotalSpent FROM Customer c JOIN Invoice i ON c.CustomerId = i.CustomerId GROUP BY c.Country ORDER BY TotalSpent DESC LIMIT 1;
================================= Tool Message =================================
Name: sql_db_query

[('USA', 523.06)]
================================== Ai Message ==================================

The country whose customers spent the most is the USA, with a total spending of 523.06.

您还可以使用LangSmith 追踪来可视化这些步骤和相关的元数据。

请注意，代理会执行多次查询，直到它获取到所需的信息：

1. 列出可用表；
2. 检索三个表的模式；
3. 通过联接操作查询多个表。

代理随后能够使用最终查询的结果生成对原始问题的答案。

代理也可以类似地处理定性问题

question = "Describe the playlisttrack table"

for step in agent_executor.stream(
    {"messages": [{"role": "user", "content": question}]},
    stream_mode="values",
):
    step["messages"][-1].pretty_print()

================================ Human Message =================================

Describe the playlisttrack table
================================== Ai Message ==================================
Tool Calls:
  sql_db_list_tables (call_fMF8eTmX5TJDJjc3Mhdg52TI)
 Call ID: call_fMF8eTmX5TJDJjc3Mhdg52TI
  Args:
================================= Tool Message =================================
Name: sql_db_list_tables

Album, Artist, Customer, Employee, Genre, Invoice, InvoiceLine, MediaType, Playlist, PlaylistTrack, Track
================================== Ai Message ==================================
Tool Calls:
  sql_db_schema (call_W8Vkk4NEodkAAIg8nexAszUH)
 Call ID: call_W8Vkk4NEodkAAIg8nexAszUH
  Args:
    table_names: PlaylistTrack
================================= Tool Message =================================
Name: sql_db_schema


CREATE TABLE "PlaylistTrack" (
	"PlaylistId" INTEGER NOT NULL, 
	"TrackId" INTEGER NOT NULL, 
	PRIMARY KEY ("PlaylistId", "TrackId"), 
	FOREIGN KEY("TrackId") REFERENCES "Track" ("TrackId"), 
	FOREIGN KEY("PlaylistId") REFERENCES "Playlist" ("PlaylistId")
)

/*
3 rows from PlaylistTrack table:
PlaylistId	TrackId
1	3402
1	3389
1	3390
*/
================================== Ai Message ==================================

The `PlaylistTrack` table is designed to associate tracks with playlists. It has the following structure:

- **PlaylistId**: An integer that serves as a foreign key referencing the `Playlist` table. It is part of the composite primary key.
- **TrackId**: An integer that serves as a foreign key referencing the `Track` table. It is also part of the composite primary key.

The primary key for this table is a composite key consisting of both `PlaylistId` and `TrackId`, ensuring that each track can be uniquely associated with a playlist. The table enforces referential integrity by linking to the `Track` and `Playlist` tables through foreign keys.

处理高基数列

为了过滤包含专有名词（如地址、歌曲名称或艺术家）的列，我们首先需要仔细检查拼写，以正确过滤数据。

我们可以通过创建一个包含数据库中所有不同专有名词的向量存储来实现这一点。然后，每当用户在问题中包含专有名词时，我们可以让代理查询该向量存储，以找到该词的正确拼写。通过这种方式，代理可以确保在构建目标查询之前，它理解用户指的是哪个实体。

首先，我们需要我们想要的每个实体的唯一值，为此我们定义一个函数，将结果解析为元素列表。

import ast
import re


def query_as_list(db, query):
    res = db.run(query)
    res = [el for sub in ast.literal_eval(res) for el in sub if el]
    res = [re.sub(r"\b\d+\b", "", string).strip() for string in res]
    return list(set(res))


artists = query_as_list(db, "SELECT Name FROM Artist")
albums = query_as_list(db, "SELECT Title FROM Album")
albums[:5]

['In Through The Out Door',
 'Transmission',
 'Battlestar Galactica (Classic), Season',
 'A Copland Celebration, Vol. I',
 'Quiet Songs']

使用此函数，我们可以创建一个检索工具，代理可以自行决定执行该工具。

让我们为这一步选择一个嵌入模型和向量存储。

选择一个嵌入模型:

选择 嵌入模型

OpenAI

Azure

Google Gemini

Google Vertex

AWS

HuggingFace

Ollama

Cohere

MistralAI

Nomic

NVIDIA

Voyage AI

IBM watsonx

Fake

pip install -qU langchain-openai

import getpass
import os

if not os.environ.get("OPENAI_API_KEY"):
  os.environ["OPENAI_API_KEY"] = getpass.getpass("Enter API key for OpenAI: ")

from langchain_openai import OpenAIEmbeddings

embeddings = OpenAIEmbeddings(model="text-embedding-3-large")

选择一个向量存储:

选择 向量存储

内存型

AstraDB

Chroma

FAISS

Milvus

MongoDB

PGVector

Pinecone

Qdrant

pip install -qU langchain-core

from langchain_core.vectorstores import InMemoryVectorStore

vector_store = InMemoryVectorStore(embeddings)

我们现在可以构建一个检索工具，用于在数据库中搜索相关的专有名词。

from langchain.agents.agent_toolkits import create_retriever_tool

_ = vector_store.add_texts(artists + albums)
retriever = vector_store.as_retriever(search_kwargs={"k": 5})
description = (
    "Use to look up values to filter on. Input is an approximate spelling "
    "of the proper noun, output is valid proper nouns. Use the noun most "
    "similar to the search."
)
retriever_tool = create_retriever_tool(
    retriever,
    name="search_proper_nouns",
    description=description,
)

API 参考：create_retriever_tool

让我们试一试

print(retriever_tool.invoke("Alice Chains"))

Alice In Chains

Alanis Morissette

Pearl Jam

Pearl Jam

Audioslave

这样，如果代理确定需要根据艺术家（例如“Alice Chains”）编写过滤器，它可以首先使用检索工具来观察列的相关值。

总结

# Add to system message
suffix = (
    "If you need to filter on a proper noun like a Name, you must ALWAYS first look up "
    "the filter value using the 'search_proper_nouns' tool! Do not try to "
    "guess at the proper name - use this function to find similar ones."
)

system = f"{system_message}\n\n{suffix}"

tools.append(retriever_tool)

agent = create_react_agent(llm, tools, prompt=system)

question = "How many albums does alis in chain have?"

for step in agent.stream(
    {"messages": [{"role": "user", "content": question}]},
    stream_mode="values",
):
    step["messages"][-1].pretty_print()

================================ Human Message =================================

How many albums does alis in chain have?
================================== Ai Message ==================================
Tool Calls:
  search_proper_nouns (call_8ryjsRPLAr79mM3Qvnq6gTOH)
 Call ID: call_8ryjsRPLAr79mM3Qvnq6gTOH
  Args:
    query: alis in chain
================================= Tool Message =================================
Name: search_proper_nouns

Alice In Chains

Aisha Duo

Xis

Da Lama Ao Caos

A-Sides
================================== Ai Message ==================================
Tool Calls:
  sql_db_list_tables (call_NJjtCpU89MBMplssjn1z0xzq)
 Call ID: call_NJjtCpU89MBMplssjn1z0xzq
  Args:
  search_proper_nouns (call_1BfrueC9koSIyi4OfMu2Ao8q)
 Call ID: call_1BfrueC9koSIyi4OfMu2Ao8q
  Args:
    query: Alice In Chains
================================= Tool Message =================================
Name: search_proper_nouns

Alice In Chains

Pearl Jam

Pearl Jam

Foo Fighters

Soundgarden
================================== Ai Message ==================================
Tool Calls:
  sql_db_schema (call_Kn09w9jd9swcNzIZ1b5MlKID)
 Call ID: call_Kn09w9jd9swcNzIZ1b5MlKID
  Args:
    table_names: Album, Artist
================================= Tool Message =================================
Name: sql_db_schema


CREATE TABLE "Album" (
	"AlbumId" INTEGER NOT NULL, 
	"Title" NVARCHAR(160) NOT NULL, 
	"ArtistId" INTEGER NOT NULL, 
	PRIMARY KEY ("AlbumId"), 
	FOREIGN KEY("ArtistId") REFERENCES "Artist" ("ArtistId")
)

/*
3 rows from Album table:
AlbumId	Title	ArtistId
1	For Those About To Rock We Salute You	1
2	Balls to the Wall	2
3	Restless and Wild	2
*/


CREATE TABLE "Artist" (
	"ArtistId" INTEGER NOT NULL, 
	"Name" NVARCHAR(120), 
	PRIMARY KEY ("ArtistId")
)

/*
3 rows from Artist table:
ArtistId	Name
1	AC/DC
2	Accept
3	Aerosmith
*/
================================== Ai Message ==================================
Tool Calls:
  sql_db_query (call_WkHRiPcBoGN9bc58MIupRHKP)
 Call ID: call_WkHRiPcBoGN9bc58MIupRHKP
  Args:
    query: SELECT COUNT(*) FROM Album WHERE ArtistId = (SELECT ArtistId FROM Artist WHERE Name = 'Alice In Chains')
================================= Tool Message =================================
Name: sql_db_query

[(1,)]
================================== Ai Message ==================================

Alice In Chains has released 1 album in the database.

如我们所见，无论是在流式步骤中还是在LangSmith 追踪中，代理都使用了search_proper_nouns工具，以检查如何正确地查询数据库以获取此特定艺术家的信息。