Skip to main content
Open In Colab在 GitHub 上打开

AWS DynamoDB

Amazon AWS DynamoDB 是一个完全托管的NoSQL数据库服务,可提供快速且可预测的性能和无缝可扩展性。

此笔记本介绍了如何使用DynamoDB存储聊天消息历史记录DynamoDBChatMessageHistory类。

设置

首先,确保您已正确配置 AWS CLI。然后确保您已安装langchain-community包,所以我们需要安装它。我们还需要安装boto3包。

pip install -U langchain-community boto3

设置 LangSmith 以实现一流的可观测性也很有帮助(但不是必需的)

# os.environ["LANGSMITH_TRACING"] = "true"
# os.environ["LANGSMITH_API_KEY"] = getpass.getpass()
from langchain_community.chat_message_histories import (
DynamoDBChatMessageHistory,
)

创建表

现在,创建DynamoDB我们将存储消息的表:

import boto3

# Get the service resource.
dynamodb = boto3.resource("dynamodb")

# Create the DynamoDB table.
table = dynamodb.create_table(
TableName="SessionTable",
KeySchema=[{"AttributeName": "SessionId", "KeyType": "HASH"}],
AttributeDefinitions=[{"AttributeName": "SessionId", "AttributeType": "S"}],
BillingMode="PAY_PER_REQUEST",
)

# Wait until the table exists.
table.meta.client.get_waiter("table_exists").wait(TableName="SessionTable")

# Print out some data about the table.
print(table.item_count)
0

DynamoDBChatMessageHistory

history = DynamoDBChatMessageHistory(table_name="SessionTable", session_id="0")

history.add_user_message("hi!")

history.add_ai_message("whats up?")
history.messages
[HumanMessage(content='hi!'), AIMessage(content='whats up?')]

具有自定义终端节点 URL 的 DynamoDBChatMessageHistory

有时,指定要连接到的 AWS 终端节点的 URL 非常有用。例如,当您在本地针对 Localstack 运行时。对于这些情况,您可以通过endpoint_url参数。

history = DynamoDBChatMessageHistory(
table_name="SessionTable",
session_id="0",
endpoint_url="http://localhost.localstack.cloud:4566",
)

具有组合键的 DynamoDBChatMessageHistory

DynamoDBChatMessageHistory 的默认键为{"SessionId": self.session_id},但您可以修改此名称以匹配您的表设计。

主键名称

您可以通过在构造函数中传入 primary_key_name 值来修改主键,从而产生以下结果:{self.primary_key_name: self.session_id}

组合键

使用现有 DynamoDB 表时,您可能需要将键结构从默认 of 修改为包括 Sort Key 在内的内容。为此,您可以使用key参数。

为 key 传递值将覆盖 primary_key 参数,生成的 key 结构将是传递的值。

composite_table = dynamodb.create_table(
TableName="CompositeTable",
KeySchema=[
{"AttributeName": "PK", "KeyType": "HASH"},
{"AttributeName": "SK", "KeyType": "RANGE"},
],
AttributeDefinitions=[
{"AttributeName": "PK", "AttributeType": "S"},
{"AttributeName": "SK", "AttributeType": "S"},
],
BillingMode="PAY_PER_REQUEST",
)

# Wait until the table exists.
composite_table.meta.client.get_waiter("table_exists").wait(TableName="CompositeTable")

# Print out some data about the table.
print(composite_table.item_count)
0
my_key = {
"PK": "session_id::0",
"SK": "langchain_history",
}

composite_key_history = DynamoDBChatMessageHistory(
table_name="CompositeTable",
session_id="0",
endpoint_url="http://localhost.localstack.cloud:4566",
key=my_key,
)

composite_key_history.add_user_message("hello, composite dynamodb table!")

composite_key_history.messages
[HumanMessage(content='hello, composite dynamodb table!')]

链接

我们可以轻松地将此消息历史类与 LCEL Runnables 结合使用

为此,我们将需要使用 OpenAI,因此我们需要安装它

from langchain_core.prompts import ChatPromptTemplate, MessagesPlaceholder
from langchain_core.runnables.history import RunnableWithMessageHistory
from langchain_openai import ChatOpenAI
prompt = ChatPromptTemplate.from_messages(
[
("system", "You are a helpful assistant."),
MessagesPlaceholder(variable_name="history"),
("human", "{question}"),
]
)

chain = prompt | ChatOpenAI()
chain_with_history = RunnableWithMessageHistory(
chain,
lambda session_id: DynamoDBChatMessageHistory(
table_name="SessionTable", session_id=session_id
),
input_messages_key="question",
history_messages_key="history",
)
# This is where we configure the session id
config = {"configurable": {"session_id": "<SESSION_ID>"}}
chain_with_history.invoke({"question": "Hi! I'm bob"}, config=config)
AIMessage(content='Hello Bob! How can I assist you today?')
chain_with_history.invoke({"question": "Whats my name"}, config=config)
AIMessage(content='Your name is Bob! Is there anything specific you would like assistance with, Bob?')