聊天模型是语言模型的变体。虽然聊天模型在底层使用语言模型，但它们公开的界面有点不同。他们没有公开“文本输入，文本输出”API，而是公开了一个接口，其中“聊天消息”是输入和输出。

聊天模型 API 相当新，所以我们仍在寻找正确的抽象。

提供了以下文档部分：

• 入门：LangChain LLM 课程提供的所有功能的概述。
• 操作指南：操作指南的集合。这些重点介绍了如何使用我们的 LLM 课程（流媒体、异步等）实现各种目标。
• 集成：关于如何将不同的 LLM 提供商与 LangChain（OpenAI、Hugging Face 等）集成的示例集合。

开始

该笔记本介绍了如何开始使用聊天模型。该界面基于消息而不是原始文本。

from langchain.chat_models import ChatOpenAI
from langchain import PromptTemplate, LLMChain
from langchain.prompts.chat import (
    ChatPromptTemplate,
    SystemMessagePromptTemplate,
    AIMessagePromptTemplate,
    HumanMessagePromptTemplate,
)
from langchain.schema import (
    AIMessage,
    HumanMessage,
    SystemMessage
)

chat = ChatOpenAI(temperature=0)

您可以通过将一条或多条消息传递给聊天模型来获得聊天完成。响应将是一条消息。LangChain 目前支持的消息类型有AIMessage, HumanMessage, SystemMessage, 和ChatMessage–ChatMessage接受任意角色参数。大多数时候，你只会处理HumanMessage, AIMessage, 和SystemMessage

chat([HumanMessage(content="Translate this sentence from English to French. I love programming.")])

AIMessage(content="J'aime programmer.", additional_kwargs={})

OpenAI 的聊天模型支持多条消息作为输入。有关更多信息，请参见此处。以下是向聊天模型发送系统和用户消息的示例：

messages = [
    SystemMessage(content="You are a helpful assistant that translates English to French."),
    HumanMessage(content="I love programming.")
]
chat(messages)

AIMessage(content="J'aime programmer.", additional_kwargs={})

您可以更进一步，使用 为多组消息生成补全generate。这将返回LLMResult带有附加message参数的 。

batch_messages = [
    [
        SystemMessage(content="You are a helpful assistant that translates English to French."),
        HumanMessage(content="I love programming.")
    ],
    [
        SystemMessage(content="You are a helpful assistant that translates English to French."),
        HumanMessage(content="I love artificial intelligence.")
    ],
]
result = chat.generate(batch_messages)
result

LLMResult(generations=[[ChatGeneration(text="J'aime programmer.", generation_info=None, message=AIMessage(content="J'aime programmer.", additional_kwargs={}))], [ChatGeneration(text="J'aime l'intelligence artificielle.", generation_info=None, message=AIMessage(content="J'aime l'intelligence artificielle.", additional_kwargs={}))]], llm_output={'token_usage': {'prompt_tokens': 57, 'completion_tokens': 20, 'total_tokens': 77}})

您可以从此 LLMResult 中恢复诸如令牌使用之类的内容

result.llm_output

{'token_usage': {'prompt_tokens': 57,
  'completion_tokens': 20,
  'total_tokens': 77}}

提示模板

您可以通过使用MessagePromptTemplate. 您可以ChatPromptTemplate从一个或多个构建一个MessagePromptTemplates。您可以使用ChatPromptTemplate's format_prompt– 这会返回一个PromptValue，您可以将其转换为字符串或 Message 对象，具体取决于您是要使用格式化值作为 llm 还是聊天模型的输入。

为方便起见，from_template模板上公开了一个方法。如果您要使用此模板，它会是这样的：

template="You are a helpful assistant that translates {input_language} to {output_language}."
system_message_prompt = SystemMessagePromptTemplate.from_template(template)
human_template="{text}"
human_message_prompt = HumanMessagePromptTemplate.from_template(human_template)

chat_prompt = ChatPromptTemplate.from_messages([system_message_prompt, human_message_prompt])

# get a chat completion from the formatted messages
chat(chat_prompt.format_prompt(input_language="English", output_language="French", text="I love programming.").to_messages())

AIMessage(content="J'adore la programmation.", additional_kwargs={})

如果想更直接的构造MessagePromptTemplate，可以在外面创建一个PromptTemplate，然后传入，eg：

prompt=PromptTemplate(
    template="You are a helpful assistant that translates {input_language} to {output_language}.",
    input_variables=["input_language", "output_language"],
)
system_message_prompt = SystemMessagePromptTemplate(prompt=prompt)

LLM链

您可以使用与之前非常相似的方式使用现有的 LLMChain - 提供提示和模型。

chain = LLMChain(llm=chat, prompt=chat_prompt)

chain.run(input_language="English", output_language="French", text="I love programming.")

"J'adore la programmation."

流媒体

通过回调处理支持流式ChatOpenAI处理。

from langchain.callbacks.streaming_stdout import StreamingStdOutCallbackHandler
chat = ChatOpenAI(streaming=True, callbacks=[StreamingStdOutCallbackHandler()], temperature=0)
resp = chat([HumanMessage(content="Write me a song about sparkling water.")])

Verse 1:
Bubbles rising to the top
A refreshing drink that never stops
Clear and crisp, it's pure delight
A taste that's sure to excite

Chorus:
Sparkling water, oh so fine
A drink that's always on my mind
With every sip, I feel alive
Sparkling water, you're my vibe

Verse 2:
No sugar, no calories, just pure bliss
A drink that's hard to resist
It's the perfect way to quench my thirst
A drink that always comes first

Chorus:
Sparkling water, oh so fine
A drink that's always on my mind
With every sip, I feel alive
Sparkling water, you're my vibe

Bridge:
From the mountains to the sea
Sparkling water, you're the key
To a healthy life, a happy soul
A drink that makes me feel whole

Chorus:
Sparkling water, oh so fine
A drink that's always on my mind
With every sip, I feel alive
Sparkling water, you're my vibe

Outro:
Sparkling water, you're the one
A drink that's always so much fun
I'll never let you go, my friend
Sparkling

此处的示例都针对使用聊天模型的某些“操作方法”指南。

• 如何使用少量镜头示例
• 如何流式传输响应

如何使用少量镜头示例

本笔记本介绍了如何在聊天模型中使用少量镜头示例。

对于如何最好地进行少量投篮提示，似乎没有达成一致意见。因此，我们还没有围绕这个巩固任何抽象，而是使用现有的抽象。

交替的人类/人工智能信息

进行少量镜头提示的第一种方法依赖于使用交替的人/人工智能信息。请参阅下面的示例。

from langchain.chat_models import ChatOpenAI
from langchain import PromptTemplate, LLMChain
from langchain.prompts.chat import (
    ChatPromptTemplate,
    SystemMessagePromptTemplate,
    AIMessagePromptTemplate,
    HumanMessagePromptTemplate,
)
from langchain.schema import (
    AIMessage,
    HumanMessage,
    SystemMessage
)

chat = ChatOpenAI(temperature=0)

template="You are a helpful assistant that translates english to pirate."
system_message_prompt = SystemMessagePromptTemplate.from_template(template)
example_human = HumanMessagePromptTemplate.from_template("Hi")
example_ai = AIMessagePromptTemplate.from_template("Argh me mateys")
human_template="{text}"
human_message_prompt = HumanMessagePromptTemplate.from_template(human_template)

chat_prompt = ChatPromptTemplate.from_messages([system_message_prompt, example_human, example_ai, human_message_prompt])
chain = LLMChain(llm=chat, prompt=chat_prompt)
# get a chat completion from the formatted messages
chain.run("I love programming.")

"I be lovin' programmin', me hearty!"

系统消息

OpenAI 提供了一个可选name参数，他们还建议将其与系统消息结合使用以进行少量提示。以下是如何执行此操作的示例。

template="You are a helpful assistant that translates english to pirate."
system_message_prompt = SystemMessagePromptTemplate.from_template(template)
example_human = SystemMessagePromptTemplate.from_template("Hi", additional_kwargs={"name": "example_user"})
example_ai = SystemMessagePromptTemplate.from_template("Argh me mateys", additional_kwargs={"name": "example_assistant"})
human_template="{text}"
human_message_prompt = HumanMessagePromptTemplate.from_template(human_template)

chat_prompt = ChatPromptTemplate.from_messages([system_message_prompt, example_human, example_ai, human_message_prompt])
chain = LLMChain(llm=chat, prompt=chat_prompt)
# get a chat completion from the formatted messages
chain.run("I love programming.")

"I be lovin' programmin', me hearty."

如何流式传输响应

本笔记本介绍了如何将流式处理与聊天模型结合使用。

from langchain.chat_models import ChatOpenAI
from langchain.schema import (
    HumanMessage,
)

from langchain.callbacks.streaming_stdout import StreamingStdOutCallbackHandler
chat = ChatOpenAI(streaming=True, callbacks=[StreamingStdOutCallbackHandler()], temperature=0)
resp = chat([HumanMessage(content="Write me a song about sparkling water.")])

Verse 1:
Bubbles rising to the top
A refreshing drink that never stops
Clear and crisp, it's pure delight
A taste that's sure to excite

Chorus:
Sparkling water, oh so fine
A drink that's always on my mind
With every sip, I feel alive
Sparkling water, you're my vibe

Verse 2:
No sugar, no calories, just pure bliss
A drink that's hard to resist
It's the perfect way to quench my thirst
A drink that always comes first

Chorus:
Sparkling water, oh so fine
A drink that's always on my mind
With every sip, I feel alive
Sparkling water, you're my vibe

Bridge:
From the mountains to the sea
Sparkling water, you're the key
To a healthy life, a happy soul
A drink that makes me feel whole

Chorus:
Sparkling water, oh so fine
A drink that's always on my mind
With every sip, I feel alive
Sparkling water, you're my vibe

Outro:
Sparkling water, you're the one
A drink that's always so much fun
I'll never let you go, my friend
Sparkling

集成

这里的例子都强调了如何与不同的聊天模型集成。

• 人择的
• 蔚蓝
• 开放人工智能
• PromptLayer ChatOpenAI

人择

本笔记本介绍了如何开始使用 Anthropic 聊天模型。

from langchain.chat_models import ChatAnthropic
from langchain.prompts.chat import (
    ChatPromptTemplate,
    SystemMessagePromptTemplate,
    AIMessagePromptTemplate,
    HumanMessagePromptTemplate,
)
from langchain.schema import (
    AIMessage,
    HumanMessage,
    SystemMessage
)

chat = ChatAnthropic()

messages = [
    HumanMessage(content="Translate this sentence from English to French. I love programming.")
]
chat(messages)

AIMessage(content=" J'aime programmer. ", additional_kwargs={})

ChatAnthropic还支持异步和流功能：from langchain.callbacks.manager import CallbackManager from langchain.callbacks.streaming_stdout import StreamingStdOutCallbackHandler await chat.agenerate([messages]) LLMResult(generations=[[ChatGeneration(text=" J'aime la programmation.", generation_info=None, message=AIMessage(content=" J'aime la programmation.", additional_kwargs={}))]], llm_output={}) chat = ChatAnthropic(streaming=True, verbose=True, callback_manager=CallbackManager([StreamingStdOutCallbackHandler()])) chat(messages) J'adore programmer. AIMessage(content=" J'adore programmer.", additional_kwargs={})人择

本笔记本介绍了如何开始使用 Anthropic 聊天模型。

from langchain.chat_models import ChatAnthropic
from langchain.prompts.chat import (
    ChatPromptTemplate,
    SystemMessagePromptTemplate,
    AIMessagePromptTemplate,
    HumanMessagePromptTemplate,
)
from langchain.schema import (
    AIMessage,
    HumanMessage,
    SystemMessage
)

chat = ChatAnthropic()

messages = [
    HumanMessage(content="Translate this sentence from English to French. I love programming.")
]
chat(messages)

AIMessage(content=" J'aime programmer. ", additional_kwargs={})

ChatAnthropic还支持异步和流功能：

from langchain.callbacks.manager import CallbackManager from langchain.callbacks.streaming_stdout import StreamingStdOutCallbackHandler await chat.agenerate([messages]) LLMResult(generations=[[ChatGeneration(text=" J'aime la programmation.", generation_info=None, message=AIMessage(content=" J'aime la programmation.", additional_kwargs={}))]], llm_output={}) chat = ChatAnthropic(streaming=True, verbose=True, callback_manager=CallbackManager([StreamingStdOutCallbackHandler()])) chat(messages) J'adore programmer. AIMessage(content=" J'adore programmer.", additional_kwargs={})

蔚蓝

此笔记本介绍了如何连接到 Azure 托管的 OpenAI 端点

from langchain.chat_models import AzureChatOpenAI
from langchain.schema import HumanMessage

BASE_URL = "https://${TODO}.openai.azure.com"
API_KEY = "..."
DEPLOYMENT_NAME = "chat"
model = AzureChatOpenAI(
    openai_api_base=BASE_URL,
    openai_api_version="2023-03-15-preview",
    deployment_name=DEPLOYMENT_NAME,
    openai_api_key=API_KEY,
    openai_api_type = "azure",
)

model([HumanMessage(content="Translate this sentence from English to French. I love programming.")])

AIMessage(content="

J'aime programmer.", additional_kwargs={})

开放人工智能

本笔记本介绍了如何开始使用 OpenAI 聊天模型。

from langchain.chat_models import ChatOpenAI
from langchain.prompts.chat import (
    ChatPromptTemplate,
    SystemMessagePromptTemplate,
    AIMessagePromptTemplate,
    HumanMessagePromptTemplate,
)
from langchain.schema import (
    AIMessage,
    HumanMessage,
    SystemMessage
)

chat = ChatOpenAI(temperature=0)

messages = [
    SystemMessage(content="You are a helpful assistant that translates English to French."),
    HumanMessage(content="Translate this sentence from English to French. I love programming.")
]
chat(messages)

AIMessage(content="J'aime programmer.", additional_kwargs={}, example=False)

您可以通过使用MessagePromptTemplate. 您可以ChatPromptTemplate从一个或多个构建一个MessagePromptTemplates。您可以使用ChatPromptTemplate's format_prompt– 这会返回一个PromptValue，您可以将其转换为字符串或 Message 对象，具体取决于您是要使用格式化值作为 llm 还是聊天模型的输入。

为方便起见，from_template模板上公开了一个方法。如果您要使用此模板，它会是这样的：

template="You are a helpful assistant that translates {input_language} to {output_language}."
system_message_prompt = SystemMessagePromptTemplate.from_template(template)
human_template="{text}"
human_message_prompt = HumanMessagePromptTemplate.from_template(human_template)

chat_prompt = ChatPromptTemplate.from_messages([system_message_prompt, human_message_prompt])

# get a chat completion from the formatted messages
chat(chat_prompt.format_prompt(input_language="English", output_language="French", text="I love programming.").to_messages())

AIMessage(content="J'adore la programmation.", additional_kwargs={})

PromptLayer ChatOpenAI

此示例展示了如何连接到PromptLayer以开始记录您的 ChatOpenAI 请求。

安装提示层

该promptlayer包需要将 PromptLayer 与 OpenAI 一起使用。使用 pip安装promptlayer。

pip install promptlayer

进口import os from langchain.chat_models import PromptLayerChatOpenAI from langchain.schema import HumanMessage

设置环境 API 密钥

您可以通过单击导航栏中的设置齿轮在www.promptlayer.com上创建一个 PromptLayer API 密钥。

将其设置为名为PROMPTLAYER_API_KEY.

os.environ["PROMPTLAYER_API_KEY"] = "**********"

像平常一样使用 PromptLayerOpenAI LLM

您可以选择传入pl_tags以使用 PromptLayer 的标记功能跟踪您的请求。

chat = PromptLayerChatOpenAI(pl_tags=["langchain"])
chat([HumanMessage(content="I am a cat and I want")])

AIMessage(content='to take a nap in a cozy spot. I search around for a suitable place and finally settle on a soft cushion on the window sill. I curl up into a ball and close my eyes, relishing the warmth of the sun on my fur. As I drift off to sleep, I can hear the birds chirping outside and feel the gentle breeze blowing through the window. This is the life of a contented cat.', additional_kwargs={})

上面的请求现在应该出现在您的PromptLayer 仪表板上。

使用 PromptLayer 轨道

如果您想使用任何PromptLayer 跟踪功能return_pl_id，您需要在实例化 PromptLayer LLM 时传递参数以获取请求 ID。

chat = PromptLayerChatOpenAI(return_pl_id=True)
chat_results = chat.generate([[HumanMessage(content="I am a cat and I want")]])

for res in chat_results.generations:
    pl_request_id = res[0].generation_info["pl_request_id"]
    promptlayer.track.score(request_id=pl_request_id, score=100)

使用它可以让您在 PromptLayer 仪表板中跟踪模型的性能。如果您使用的是提示模板，您也可以将模板附加到请求中。总体而言，这使您有机会在 PromptLayer 仪表板中跟踪不同模板和模型的性能。