前提条件
本指南假定您熟悉以下概念:
流式处理对于使基于 LLM 的应用程序对最终用户感觉响应迅速至关重要。
重要的 LangChain 原语如 聊天模型、输出解析器、提示、检索器 和 代理 都实现了 LangChain 可运行接口。
该接口提供了两种通用方法来流式传输内容:
- 同步
stream和异步astream:这是一个默认实现的流式传输方法,它从链中流式传输最终输出。 - 异步
astream_events和异步astream_log:这些方法提供了一种从链中流式传输中间步骤和最终输出的方法。
让我们来看看这两种方法,并尝试了解如何使用它们。
使用 Stream
所有 Runnable 对象都实现了一个名为 stream 的同步方法和一个异步变体 astream。
这些方法旨在分块流式传输最终输出,每当一个块可用时立即生成该块。
只有当程序中的所有步骤都知道如何处理输入流时,流式传输才有可能;即一次处理一个输入块,并生成相应的输出块。
这种处理的复杂性可能会有所不同,从发出 LLM 生成的标记等简单任务到在整个 JSON 完成之前流式传输 JSON 结果的部分等更具挑战性的任务。
探索流式传输的最佳起点是 LLM 应用程序中最重要的组件——LLM 本身!
LLM 和聊天模型
大型语言模型及其聊天变体是基于 LLM 的应用程序中的主要瓶颈。
大型语言模型可能需要几秒钟才能生成对查询的完整响应。这远比使应用程序对最终用户感觉响应迅速的~200-300 毫秒阈值慢得多。
使应用程序感觉更响应的关键策略是显示中间进度;即,逐字元地流式传输模型的输出。
我们将展示使用聊天模型进行流式传输的示例。从以下选项中选择一个:
OpenAI
pip install -qU langchain-openai
import getpassimport osos.environ["OPENAI_API_KEY"] = getpass.getpass()from langchain_openai import ChatOpenAImodel = ChatOpenAI(model="gpt-3.5-turbo-0125")
Anthropic
pip install -qU langchain-anthropic
import getpassimport osos.environ["ANTHROPIC_API_KEY"] = getpass.getpass()from langchain_anthropic import ChatAnthropicmodel = ChatAnthropic(model="claude-3-sonnet-20240229")
Azure
pip install -qU langchain-openai
import getpassimport osos.environ["AZURE_OPENAI_API_KEY"] = getpass.getpass()from langchain_openai import AzureChatOpenAImodel = AzureChatOpenAI(azure_endpoint=os.environ["AZURE_OPENAI_ENDPOINT"],azure_deployment=os.environ["AZURE_OPENAI_DEPLOYMENT_NAME"],openai_api_version=os.environ["AZURE_OPENAI_API_VERSION"],)
pip install -qU langchain-google-vertexai
import getpassimport osos.environ["GOOGLE_API_KEY"] = getpass.getpass()from langchain_google_vertexai import ChatVertexAImodel = ChatVertexAI(model="gemini-pro")
Cohere
pip install -qU langchain-cohere
import getpassimport osos.environ["COHERE_API_KEY"] = getpass.getpass()from langchain_cohere import ChatCoheremodel = ChatCohere(model="command-r")
FireworksAI
pip install -qU langchain-fireworks
import getpassimport osos.environ["FIREWORKS_API_KEY"] = getpass.getpass()from langchain_fireworks import ChatFireworksmodel = ChatFireworks(model="accounts/fireworks/models/mixtral-8x7b-instruct")
MistralAI
pip install -qU langchain-mistralai
import getpassimport osos.environ["MISTRAL_API_KEY"] = getpass.getpass()from langchain_mistralai import ChatMistralAImodel = ChatMistralAI(model="mistral-large-latest")
TogetherAI
pip install -qU langchain-openai
import getpassimport osos.environ["TOGETHER_API_KEY"] = getpass.getpass()from langchain_openai import ChatOpenAImodel = ChatOpenAI(base_url="https://api.together.xyz/v1",api_key=os.environ["TOGETHER_API_KEY"],model="mistralai/Mixtral-8x7B-Instruct-v0.1",)
让我们从同步 stream API 开始:
chunks = []for chunk in model.stream("what color is the sky?"):chunks.append(chunk)print(chunk.content, end="|", flush=True)
The| sky| appears| blue| during| the| day|.
如果你在异步环境中工作,可以考虑使用异步 astream API:
chunks = []async for chunk in model.astream("what color is the sky?"):chunks.append(chunk)print(chunk.content, end="|", flush=True)
The| sky| appears| blue| during| the| day|.
让我们检查其中一个块
chunks[0]
AIMessageChunk(content='The', id='run-b36bea64-5511-4d7a-b6a3-a07b3db0c8e7')
我们得到了一个称为 AIMessageChunk 的东西。这个块代表 AIMessage 的一部分。
消息块是累加设计的——可以简单地将它们加起来以获得到目前为止的响应状态!
chunks[0] + chunks[1] + chunks[2] + chunks[3] + chunks[4]
AIMessageChunk(content='The sky appears blue during', id='run-b36bea64-5511-4d7a-b6a3-a07b3db0c8e7')
链
几乎所有 LLM 应用程序都涉及的不仅仅是调用语言模型的步骤。
让我们使用 LangChain 表达式语言 (LCEL) 构建一个简单的链,该链将提示、模型和解析器结合起来,并验证流式传输是否有效。
我们将使用 StrOutputParser 来解析模型的输出。这是一个简单的解析器,它从 AIMessageChunk 中提取 content 字段,给我们模型返回的 token。
小贴士
LCEL 是一种通过将不同的 LangChain 原语连接在一起来指定“程序”的声明性方式。使用 LCEL 创建的链受益于 stream 和 astream 的自动实现,允许流式传输最终输出。事实上,使用 LCEL 创建的链实现了整个标准的可运行接口。
from langchain_core.output_parsers import StrOutputParserfrom langchain_core.prompts import ChatPromptTemplateprompt = ChatPromptTemplate.from_template("tell me a joke about {topic}")parser = StrOutputParser()chain = prompt | model | parserasync for chunk in chain.astream({"topic": "parrot"}):print(chunk, end="|", flush=True)
API 参考:StrOutputParser | ChatPromptTemplate
Here|'s| a| joke| about| a| par|rot|:|A man| goes| to| a| pet| shop| to| buy| a| par|rot|.| The| shop| owner| shows| him| two| stunning| pa|rr|ots| with| beautiful| pl|um|age|.|"|There|'s| a| talking| par|rot| an|d a| non|-|talking| par|rot|,"| the| owner| says|.| "|The| talking| par|rot| costs| $|100|,| an|d the| non|-|talking| par|rot| is| $|20|."|The| man| says|,| "|I|'ll| take| the| non|-|talking| par|rot| at| $|20|."|He| pays| an|d leaves| with| the| par|rot|.| As| he|'s| walking| down| the| street|,| the| par|rot| looks| up| at| him| an|d says|,| "|You| know|,| you| really| are| a| stupi|d man|!"|The| man| is| stun|ne|d an|d looks| at| the| par|rot| in| dis|bel|ief|.| The| par|rot| continues|,| "|Yes|,| you| got| r|ippe|d off| big| time|!| I| can| talk| just| as| well| as| that| other| par|rot|,| an|d you| only| pai|d $|20| |for| me|!"|
请注意,即使我们在上述链的末尾使用 parser,我们仍然得到了流式输出。parser 对每个流式块单独操作。许多 LCEL 原语 也支持这种变换式的传递流式处理,这在构建应用程序时非常方便。
自定义函数可以 设计为返回生成器,这些生成器能够对流进行操作。
某些可运行的对象,如 提示模板 和 聊天模型,无法处理单个块,而是会聚合所有先前的步骤。此类可运行对象可能会中断流式处理过程。
注意
LangChain 表达式语言允许你将链的构建与其使用模式(例如,同步/异步,批处理/流式处理等)分开。如果这与你正在构建的内容无关,你也可以依靠标准的命令式编程方法,分别调用每个组件上的 invoke、batch 或 stream,将结果分配给变量,然后根据需要在下游使用它们。
使用输入流
如果你想在生成输出时流式传输 JSON,该怎么办?
如果依赖 json.loads 来解析部分 JSON,解析将失败,因为部分 JSON 不是有效的 JSON。
你可能会完全不知道该怎么办,并声称无法流式传输 JSON。
实际上是有办法的——解析器需要对输入流进行操作,并尝试将部分 JSON“自动完成”为有效状态。
让我们看看这样的解析器在实际中的应用,以理解这意味着什么。
import jsondef json_stream_parser(input_stream):buffer = ""for chunk in input_stream:buffer += chunktry:data = json.loads(buffer)yield databuffer = ""except json.JSONDecodeError:continue# 示例使用input_stream = ['{"key1": "val', 'ue1", "}", '{"key2":', ' "value2"}']for parsed_json in json_stream_parser(input_stream):print(parsed_json)
这个示例展示了如何流式解析 JSON 输出。在这个例子中,json_stream_parser 函数接受一个输入流,将其缓冲并尝试解析 JSON。如果解析失败(因为 JSON 仍然不完整),它将继续累积更多的块,直到可以成功解析为止。
这种方法使得即使在输出 JSON 尚未完全生成时,也可以流式传输和解析 JSON。
from langchain_core.output_parsers import JsonOutputParserchain = (model | JsonOutputParser()) # Due to a bug in older versions of Langchain, JsonOutputParser did not stream results from some modelsasync for text in chain.astream("output a list of the countries france, spain and japan and their populations in JSON format. "'Use a dict with an outer key of "countries" which contains a list of countries. '"Each country should have the key `name` and `population`"):print(text, flush=True)
API 参考:JsonOutputParser
{}{'countries': []}{'countries': [{}]}{'countries': [{'name': ''}]}{'countries': [{'name': 'France'}]}{'countries': [{'name': 'France', 'population': 67}]}{'countries': [{'name': 'France', 'population': 67413}]}{'countries': [{'name': 'France', 'population': 67413000}]}{'countries': [{'name': 'France', 'population': 67413000}, {}]}{'countries': [{'name': 'France', 'population': 67413000}, {'name': ''}]}{'countries': [{'name': 'France', 'population': 67413000}, {'name': 'Spain'}]}{'countries': [{'name': 'France', 'population': 67413000}, {'name': 'Spain', 'population': 47}]}{'countries': [{'name': 'France', 'population': 67413000}, {'name': 'Spain', 'population': 47351}]}{'countries': [{'name': 'France', 'population': 67413000}, {'name': 'Spain', 'population': 47351567}]}{'countries': [{'name': 'France', 'population': 67413000}, {'name': 'Spain', 'population': 47351567}, {}]}{'countries': [{'name': 'France', 'population': 67413000}, {'name': 'Spain', 'population': 47351567}, {'name': ''}]}{'countries': [{'name': 'France', 'population': 67413000}, {'name': 'Spain', 'population': 47351567}, {'name': 'Japan'}]}{'countries': [{'name': 'France', 'population': 67413000}, {'name': 'Spain', 'population': 47351567}, {'name': 'Japan', 'population': 125}]}{'countries': [{'name': 'France', 'population': 67413000}, {'name': 'Spain', 'population': 47351567}, {'name': 'Japan', 'population': 125584}]}{'countries': [{'name': 'France', 'population': 67413000}, {'name': 'Spain', 'population': 47351567}, {'name': 'Japan', 'population': 125584000}]}
现在,让我们破坏流式传输。我们将使用之前的示例,并在末尾附加一个提取函数,从最终的 JSON 中提取国家名称。
危险
在链中任何操作最终输入而不是输入流的步骤,都可能通过 stream 或 astream 破坏流式功能。
提示
稍后,我们将讨论 astream_events API,该 API 从中间步骤流式传输结果。即使链中包含仅操作最终输入的步骤,该 API 也会从中间步骤流式传输结果。
以下是如何破坏流式传输的示例:
import jsondef json_stream_parser(input_stream):buffer = ""for chunk in input_stream:buffer += chunktry:data = json.loads(buffer)yield databuffer = ""except json.JSONDecodeError:continuedef extract_countries(json_data):if "countries" in json_data:return json_data["countries"]return []# 示例使用input_stream = ['{"countries": ["USA",', '"Canada", "Mexico"]}', '{"cities": ["New York",', ' "Toronto"]}']# 这是流式解析 JSONfor parsed_json in json_stream_parser(input_stream):print(parsed_json)# 这是从最终 JSON 中提取国家的步骤for parsed_json in json_stream_parser(input_stream):countries = extract_countries(parsed_json)print(countries)
在这个示例中,json_stream_parser 函数继续逐块解析 JSON 数据。然而,extract_countries 函数则需要完整的 JSON 数据来提取国家名称。这会破坏流式传输的功能,因为提取国家的操作需要等待整个 JSON 完成后才能执行。
如果我们尝试流式传输每个步骤的中间结果,则需要使用更高级的 API,如 astream_events。以下是如何实现的示例:
async def astream_events(chain, input):events = []async for event in chain.astream_events(input):events.append(event)print(event)return events# 示例使用import asyncioinput_data = 'some input data for chain'async def main():events = await astream_events(chain, input_data)for event in events:print(event)asyncio.run(main())
通过这种方式,即使链中包含仅操作最终输入的步骤,也可以从中间步骤流式传输结果。
from langchain_core.output_parsers import (JsonOutputParser,)# A function that operates on finalized inputs# rather than on an input_streamdef _extract_country_names(inputs):"""A function that does not operates on input streams and breaks streaming."""if not isinstance(inputs, dict):return ""if "countries" not in inputs:return ""countries = inputs["countries"]if not isinstance(countries, list):return ""country_names = [country.get("name") for country in countries if isinstance(country, dict)]return country_nameschain = model | JsonOutputParser() | _extract_country_namesasync for text in chain.astream("output a list of the countries france, spain and japan and their populations in JSON format. "'Use a dict with an outer key of "countries" which contains a list of countries. '"Each country should have the key `name` and `population`"):print(text, end="|", flush=True)
API 参考: JsonOutputParser
['France', 'Spain', 'Japan']|
生成器函数
让我们使用可以对输入流进行操作的生成器函数来修复流式传输。
提示
生成器函数(使用 yield 的函数)允许编写对输入流进行操作的代码。
下面是如何实现这一点的示例:
import json# 解析JSON流的生成器函数def json_stream_parser(input_stream):buffer = ""for chunk in input_stream:buffer += chunktry:data = json.loads(buffer)yield databuffer = ""except json.JSONDecodeError:continue# 提取国家名称的生成器函数def extract_countries_from_stream(input_stream):for json_data in json_stream_parser(input_stream):if "countries" in json_data:yield json_data["countries"]else:yield []# 示例使用input_stream = ['{"countries": ["France",', '"Spain", "Japan"]}', '{"cities": ["Paris",', ' "Madrid"]}']for countries in extract_countries_from_stream(input_stream):print(countries)
在这个示例中,json_stream_parser 生成器函数解析输入流中的 JSON 数据块,并生成已解析的 JSON 数据。extract_countries_from_stream 生成器函数接收这些已解析的 JSON 数据块,并从中提取国家名称。
这样,提取国家的操作可以逐块进行,而不是等待整个 JSON 完成,从而保持了流式传输的功能。
from langchain_core.output_parsers import JsonOutputParserasync def _extract_country_names_streaming(input_stream):"""A function that operates on input streams."""country_names_so_far = set()async for input in input_stream:if not isinstance(input, dict):continueif "countries" not in input:continuecountries = input["countries"]if not isinstance(countries, list):continuefor country in countries:name = country.get("name")if not name:continueif name not in country_names_so_far:yield namecountry_names_so_far.add(name)chain = model | JsonOutputParser() | _extract_country_names_streamingasync for text in chain.astream("output a list of the countries france, spain and japan and their populations in JSON format. "'Use a dict with an outer key of "countries" which contains a list of countries. '"Each country should have the key `name` and `population`",):print(text, end="|", flush=True)
API 参考: JsonOutputParser
France|Spain|Japan|
注意
由于上述代码依赖于 JSON 自动完成,你可能会看到国家名称的部分名称(例如
Sp和Spain),这并不是我们想要的提取结果!我们专注于流式传输的概念,而不一定是链的结果。
非流式组件
某些内置组件如 Retrievers 不提供任何 streaming 功能。如果我们尝试对它们进行 streaming 会发生什么? 🤨
在这种情况下,流式传输将会中断,因为这些组件需要完整的数据才能进行处理。让我们看看一个例子:
# 示例非流式组件class NonStreamingRetriever:def retrieve(self, query):# 模拟非流式的检索操作return ["Document 1", "Document 2", "Document 3"]# 流式处理链中的非流式组件def process_with_retriever(input_stream, retriever):for chunk in input_stream:# 非流式检索器在这里中断流式传输results = retriever.retrieve(chunk)for result in results:yield result# 示例使用input_stream = ["query1", "query2", "query3"]retriever = NonStreamingRetriever()for result in process_with_retriever(input_stream, retriever):print(result)
在这个示例中,NonStreamingRetriever 组件需要完整的查询才能进行检索操作,并返回结果。这会中断流式传输,因为它不能逐块处理输入。
为了解决这个问题,可以将非流式组件与流式组件隔离开来,并确保仅在整个输入流完成后调用非流式组件的操作。这可以通过在链中使用中间步骤来实现。
但是,如果我们确实需要从非流式组件中获取中间步骤的结果,我们可以使用 astream_events API,如下所示:
async def astream_events(chain, input):events = []async for event in chain.astream_events(input):events.append(event)print(event)return events# 示例使用import asyncioinput_data = 'some input data for chain'async def main():events = await astream_events(chain, input_data)for event in events:print(event)asyncio.run(main())
这样,即使链中包含仅操作最终输入的步骤,astream_events 也可以从中间步骤流式传输结果。
from langchain_community.vectorstores import FAISSfrom langchain_core.output_parsers import StrOutputParserfrom langchain_core.prompts import ChatPromptTemplatefrom langchain_core.runnables import RunnablePassthroughfrom langchain_openai import OpenAIEmbeddingstemplate = """Answer the question based only on the following context:{context}Question: {question}"""prompt = ChatPromptTemplate.from_template(template)vectorstore = FAISS.from_texts(["harrison worked at kensho", "harrison likes spicy food"],embedding=OpenAIEmbeddings(),)retriever = vectorstore.as_retriever()chunks = [chunk for chunk in retriever.stream("where did harrison work?")]chunks
API 参考: FAISS | StrOutputParser | ChatPromptTemplate | RunnablePassthrough | OpenAIEmbeddings
[[Document(page_content='harrison worked at kensho'), Document(page_content='harrison likes spicy food')]]
流式传输刚刚从该组件中产生了最终结果。
这很好 🥹!并非所有组件都必须实现流式传输——在某些情况下,流式传输要么是不必要的,要么很困难,要么根本没有意义。
提示
使用非流式组件构建的 LCEL 链,在许多情况下仍然可以流式传输,流式传输的部分输出将从链中的最后一个非流式步骤之后开始。
retrieval_chain = ({"context": retriever.with_config(run_name="Docs"),"question": RunnablePassthrough(),}| prompt| model| StrOutputParser())
for chunk in retrieval_chain.stream("Where did harrison work? " "Write 3 made up sentences about this place."):print(chunk, end="|", flush=True)
Base|d on| the| given| context|,| Harrison| worke|d at| K|ens|ho|.|Here| are| |3| |made| up| sentences| about| this| place|:|1|.| K|ens|ho| was| a| cutting|-|edge| technology| company| known| for| its| innovative| solutions| in| artificial| intelligence| an|d data| analytics|.|2|.| The| modern| office| space| at| K|ens|ho| feature|d open| floor| plans|,| collaborative| work|sp|aces|,| an|d a| vib|rant| atmosphere| that| fos|tere|d creativity| an|d team|work|.|3|.| With| its| prime| location| in| the| heart| of| the| city|,| K|ens|ho| attracte|d top| talent| from| aroun|d the| worl|d,| creating| a| diverse| an|d dynamic| work| environment|.|
使用流式事件
事件流是一个beta API。这个 API 可能会根据反馈略有变化。
注意
本指南演示了 V2 API,并需要 langchain-core >= 0.2。对于与旧版本 LangChain 兼容的 V1 API,请参阅这里。
import langchain_coreprint(langchain_core.__version__)
为了使 astream_events API 正常工作:
- 在代码中尽可能使用
async(例如,异步工具等) - 如果定义自定义函数/可运行时,请传递回调函数
- 每当在没有 LCEL 的情况下使用可运行组件时,请确保在 LLM 上调用
.astream()而不是.ainvoke,以强制 LLM 流式传输令牌。 - 如果有任何问题,请告诉我们! :)
事件参考
下面是一个参考表,显示各种可运行对象可能发出的一些事件。
注意
当流式传输正确实现时,可运行对象的输入将直到完全消耗输入流后才会知道。这意味着 inputs 通常仅包含在 end 事件中,而不是在 start 事件中。
| 事件 | 名称 | 块 | 输入 | 输出 |
|---|---|---|---|---|
| on_chat_model_start | [model name] | {“messages”: [[SystemMessage, HumanMessage]]} | ||
| on_chat_model_stream | [model name] | AIMessageChunk(content=”hello”) | ||
| on_chat_model_end | [model name] | {“messages”: [[SystemMessage, HumanMessage]]} | AIMessageChunk(content=”hello world”) | |
| on_llm_start | [model name] | {‘input’: ‘hello’} | ||
| on_llm_stream | [model name] | ‘Hello’ | ||
| on_llm_end | [model name] | ‘Hello human!’ | ||
| on_chain_start | format_docs | |||
| on_chain_stream | format_docs | “hello world!, goodbye world!” | ||
| on_chain_end | format_docs | [Document(…)] | “hello world!, goodbye world!” | |
| on_tool_start | some_tool | {“x”: 1, “y”: “2”} | ||
| on_tool_end | some_tool | {“x”: 1, “y”: “2”} | ||
| on_retriever_start | [retriever name] | {“query”: “hello”} | ||
| on_retriever_end | [retriever name] | {“query”: “hello”} | [Document(…), ..] | |
| on_prompt_start | [template_name] | {“question”: “hello”} | ||
| on_prompt_end | [template_name] | {“question”: “hello”} | ChatPromptValue(messages: [SystemMessage, …]) |
聊天模型
让我们首先看一下聊天模型产生的事件。
events = []async for event in model.astream_events("hello", version="v2"):events.append(event)
/home/eugene/src/langchain/libs/core/langchain_core/_api/beta_decorator.py:87: LangChainBetaWarning: This API is in beta and may change in the future.warn_beta(
备注
嘿,API 中那个有趣的 version=”v2” 参数是什么意思?😾
这是一个 beta API,我们几乎肯定会对它进行一些修改(事实上,我们已经修改了!)。
这个版本参数可以让我们尽量减少对您代码的破坏性修改。
简而言之,我们现在让你烦恼,以后就不用再让你烦恼了。
v2 “仅适用于 langchain-core>=0.2.0。
让我们来看看开始事件和结束事件中的几个事件。
events[:3]
[{'event': 'on_chat_model_start','data': {'input': 'hello'},'name': 'ChatAnthropic','tags': [],'run_id': 'a81e4c0f-fc36-4d33-93bc-1ac25b9bb2c3','metadata': {}},{'event': 'on_chat_model_stream','data': {'chunk': AIMessageChunk(content='Hello', id='run-a81e4c0f-fc36-4d33-93bc-1ac25b9bb2c3')},'run_id': 'a81e4c0f-fc36-4d33-93bc-1ac25b9bb2c3','name': 'ChatAnthropic','tags': [],'metadata': {}},{'event': 'on_chat_model_stream','data': {'chunk': AIMessageChunk(content='!', id='run-a81e4c0f-fc36-4d33-93bc-1ac25b9bb2c3')},'run_id': 'a81e4c0f-fc36-4d33-93bc-1ac25b9bb2c3','name': 'ChatAnthropic','tags': [],'metadata': {}}]
events[-2:]
[{'event': 'on_chat_model_stream','data': {'chunk': AIMessageChunk(content='?', id='run-a81e4c0f-fc36-4d33-93bc-1ac25b9bb2c3')},'run_id': 'a81e4c0f-fc36-4d33-93bc-1ac25b9bb2c3','name': 'ChatAnthropic','tags': [],'metadata': {}},{'event': 'on_chat_model_end','data': {'output': AIMessageChunk(content='Hello! How can I assist you today?', id='run-a81e4c0f-fc36-4d33-93bc-1ac25b9bb2c3')},'run_id': 'a81e4c0f-fc36-4d33-93bc-1ac25b9bb2c3','name': 'ChatAnthropic','tags': [],'metadata': {}}]
链
让我们重新审视一下解析流式JSON的示例链,以探索流式事件API。
chain = (model | JsonOutputParser()) # Due to a bug in older versions of Langchain, JsonOutputParser did not stream results from some modelsevents = [eventasync for event in chain.astream_events("output a list of the countries france, spain and japan and their populations in JSON format. "'Use a dict with an outer key of "countries" which contains a list of countries. '"Each country should have the key `name` and `population`",version="v2",)]
如果你检查前几个事件,你会注意到有3个不同的开始事件,而不是2个开始事件。
这三个开始事件对应于:
- 链(模型 + 解析器)
- 模型
- 解析器
events[:3]
[{'event': 'on_chain_start','data': {'input': 'output a list of the countries france, spain and japan and their populations in JSON format. Use a dict with an outer key of "countries" which contains a list of countries. Each country should have the key `name` and `population`'},'name': 'RunnableSequence','tags': [],'run_id': '4765006b-16e2-4b1d-a523-edd9fd64cb92','metadata': {}},{'event': 'on_chat_model_start','data': {'input': {'messages': [[HumanMessage(content='output a list of the countries france, spain and japan and their populations in JSON format. Use a dict with an outer key of "countries" which contains a list of countries. Each country should have the key `name` and `population`')]]}},'name': 'ChatAnthropic','tags': ['seq:step:1'],'run_id': '0320c234-7b52-4a14-ae4e-5f100949e589','metadata': {}},{'event': 'on_chat_model_stream','data': {'chunk': AIMessageChunk(content='{', id='run-0320c234-7b52-4a14-ae4e-5f100949e589')},'run_id': '0320c234-7b52-4a14-ae4e-5f100949e589','name': 'ChatAnthropic','tags': ['seq:step:1'],'metadata': {}}]
如果你查看最后的3个事件,你会看到什么?中间的事件呢?
让我们使用这个API来输出模型和解析器的流事件。我们忽略开始事件、结束事件以及来自链的事件。
num_events = 0async for event in chain.astream_events("output a list of the countries france, spain and japan and their populations in JSON format. "'Use a dict with an outer key of "countries" which contains a list of countries. '"Each country should have the key `name` and `population`",version="v2",):kind = event["event"]if kind == "on_chat_model_stream":print(f"Chat model chunk: {repr(event['data']['chunk'].content)}",flush=True,)if kind == "on_parser_stream":print(f"Parser chunk: {event['data']['chunk']}", flush=True)num_events += 1if num_events > 30:# Truncate the outputprint("...")break
Chat model chunk: '{'Parser chunk: {}Chat model chunk: '\n 'Chat model chunk: '"'Chat model chunk: 'countries'Chat model chunk: '":'Chat model chunk: ' ['Parser chunk: {'countries': []}Chat model chunk: '\n 'Chat model chunk: '{'Parser chunk: {'countries': [{}]}Chat model chunk: '\n 'Chat model chunk: '"'Chat model chunk: 'name'Chat model chunk: '":'Chat model chunk: ' "'Parser chunk: {'countries': [{'name': ''}]}Chat model chunk: 'France'Parser chunk: {'countries': [{'name': 'France'}]}Chat model chunk: '",'Chat model chunk: '\n 'Chat model chunk: '"'Chat model chunk: 'population'...
因为模型和解析器都支持流式传输,我们可以看到两个组件的实时流事件!很酷,不是吗?🦜
过滤事件
由于这个API产生了大量的事件,能够对事件进行过滤是很有用的。
你可以通过组件的名称、组件的标签或者组件的类型来过滤。
按名称过滤
chain = model.with_config({"run_name": "model"}) | JsonOutputParser().with_config({"run_name": "my_parser"})max_events = 0async for event in chain.astream_events("output a list of the countries france, spain and japan and their populations in JSON format. "'Use a dict with an outer key of "countries" which contains a list of countries. '"Each country should have the key `name` and `population`",version="v2",include_names=["my_parser"],):print(event)max_events += 1if max_events > 10:# Truncate outputprint("...")break
{'event': 'on_parser_start', 'data': {'input': 'output a list of the countries france, spain and japan and their populations in JSON format. Use a dict with an outer key of "countries" which contains a list of countries. Each country should have the key `name` and `population`'}, 'name': 'my_parser', 'tags': ['seq:step:2'], 'run_id': 'e058d750-f2c2-40f6-aa61-10f84cd671a9', 'metadata': {}}{'event': 'on_parser_stream', 'data': {'chunk': {}}, 'run_id': 'e058d750-f2c2-40f6-aa61-10f84cd671a9', 'name': 'my_parser', 'tags': ['seq:step:2'], 'metadata': {}}{'event': 'on_parser_stream', 'data': {'chunk': {'countries': []}}, 'run_id': 'e058d750-f2c2-40f6-aa61-10f84cd671a9', 'name': 'my_parser', 'tags': ['seq:step:2'], 'metadata': {}}{'event': 'on_parser_stream', 'data': {'chunk': {'countries': [{}]}}, 'run_id': 'e058d750-f2c2-40f6-aa61-10f84cd671a9', 'name': 'my_parser', 'tags': ['seq:step:2'], 'metadata': {}}{'event': 'on_parser_stream', 'data': {'chunk': {'countries': [{'name': ''}]}}, 'run_id': 'e058d750-f2c2-40f6-aa61-10f84cd671a9', 'name': 'my_parser', 'tags': ['seq:step:2'], 'metadata': {}}{'event': 'on_parser_stream', 'data': {'chunk': {'countries': [{'name': 'France'}]}}, 'run_id': 'e058d750-f2c2-40f6-aa61-10f84cd671a9', 'name': 'my_parser', 'tags': ['seq:step:2'], 'metadata': {}}{'event': 'on_parser_stream', 'data': {'chunk': {'countries': [{'name': 'France', 'population': 67}]}}, 'run_id': 'e058d750-f2c2-40f6-aa61-10f84cd671a9', 'name': 'my_parser', 'tags': ['seq:step:2'], 'metadata': {}}{'event': 'on_parser_stream', 'data': {'chunk': {'countries': [{'name': 'France', 'population': 67413}]}}, 'run_id': 'e058d750-f2c2-40f6-aa61-10f84cd671a9', 'name': 'my_parser', 'tags': ['seq:step:2'], 'metadata': {}}{'event': 'on_parser_stream', 'data': {'chunk': {'countries': [{'name': 'France', 'population': 67413000}]}}, 'run_id': 'e058d750-f2c2-40f6-aa61-10f84cd671a9', 'name': 'my_parser', 'tags': ['seq:step:2'], 'metadata': {}}{'event': 'on_parser_stream', 'data': {'chunk': {'countries': [{'name': 'France', 'population': 67413000}, {}]}}, 'run_id': 'e058d750-f2c2-40f6-aa61-10f84cd671a9', 'name': 'my_parser', 'tags': ['seq:step:2'], 'metadata': {}}{'event': 'on_parser_stream', 'data': {'chunk': {'countries': [{'name': 'France', 'population': 67413000}, {'name': ''}]}}, 'run_id': 'e058d750-f2c2-40f6-aa61-10f84cd671a9', 'name': 'my_parser', 'tags': ['seq:step:2'], 'metadata': {}}...
根据类型
chain = model.with_config({"run_name": "model"}) | JsonOutputParser().with_config({"run_name": "my_parser"})max_events = 0async for event in chain.astream_events('output a list of the countries france, spain and japan and their populations in JSON format. Use a dict with an outer key of "countries" which contains a list of countries. Each country should have the key `name` and `population`',version="v2",include_types=["chat_model"],):print(event)max_events += 1if max_events > 10:# Truncate outputprint("...")break
{'event': 'on_chat_model_start', 'data': {'input': 'output a list of the countries france, spain and japan and their populations in JSON format. Use a dict with an outer key of "countries" which contains a list of countries. Each country should have the key `name` and `population`'}, 'name': 'model', 'tags': ['seq:step:1'], 'run_id': 'db246792-2a91-4eb3-a14b-29658947065d', 'metadata': {}}{'event': 'on_chat_model_stream', 'data': {'chunk': AIMessageChunk(content='{', id='run-db246792-2a91-4eb3-a14b-29658947065d')}, 'run_id': 'db246792-2a91-4eb3-a14b-29658947065d', 'name': 'model', 'tags': ['seq:step:1'], 'metadata': {}}{'event': 'on_chat_model_stream', 'data': {'chunk': AIMessageChunk(content='\n ', id='run-db246792-2a91-4eb3-a14b-29658947065d')}, 'run_id': 'db246792-2a91-4eb3-a14b-29658947065d', 'name': 'model', 'tags': ['seq:step:1'], 'metadata': {}}{'event': 'on_chat_model_stream', 'data': {'chunk': AIMessageChunk(content='"', id='run-db246792-2a91-4eb3-a14b-29658947065d')}, 'run_id': 'db246792-2a91-4eb3-a14b-29658947065d', 'name': 'model', 'tags': ['seq:step:1'], 'metadata': {}}{'event': 'on_chat_model_stream', 'data': {'chunk': AIMessageChunk(content='countries', id='run-db246792-2a91-4eb3-a14b-29658947065d')}, 'run_id': 'db246792-2a91-4eb3-a14b-29658947065d', 'name': 'model', 'tags': ['seq:step:1'], 'metadata': {}}{'event': 'on_chat_model_stream', 'data': {'chunk': AIMessageChunk(content='":', id='run-db246792-2a91-4eb3-a14b-29658947065d')}, 'run_id': 'db246792-2a91-4eb3-a14b-29658947065d', 'name': 'model', 'tags': ['seq:step:1'], 'metadata': {}}{'event': 'on_chat_model_stream', 'data': {'chunk': AIMessageChunk(content=' [', id='run-db246792-2a91-4eb3-a14b-29658947065d')}, 'run_id': 'db246792-2a91-4eb3-a14b-29658947065d', 'name': 'model', 'tags': ['seq:step:1'], 'metadata': {}}{'event': 'on_chat_model_stream', 'data': {'chunk': AIMessageChunk(content='\n ', id='run-db246792-2a91-4eb3-a14b-29658947065d')}, 'run_id': 'db246792-2a91-4eb3-a14b-29658947065d', 'name': 'model', 'tags': ['seq:step:1'], 'metadata': {}}{'event': 'on_chat_model_stream', 'data': {'chunk': AIMessageChunk(content='{', id='run-db246792-2a91-4eb3-a14b-29658947065d')}, 'run_id': 'db246792-2a91-4eb3-a14b-29658947065d', 'name': 'model', 'tags': ['seq:step:1'], 'metadata': {}}{'event': 'on_chat_model_stream', 'data': {'chunk': AIMessageChunk(content='\n ', id='run-db246792-2a91-4eb3-a14b-29658947065d')}, 'run_id': 'db246792-2a91-4eb3-a14b-29658947065d', 'name': 'model', 'tags': ['seq:step:1'], 'metadata': {}}{'event': 'on_chat_model_stream', 'data': {'chunk': AIMessageChunk(content='"', id='run-db246792-2a91-4eb3-a14b-29658947065d')}, 'run_id': 'db246792-2a91-4eb3-a14b-29658947065d', 'name': 'model', 'tags': ['seq:step:1'], 'metadata': {}}...
按标签过滤
注意
标签是由给定可运行组件的子组件继承的。
如果你使用标签来过滤,请确保这是你想要的。
chain = (model | JsonOutputParser()).with_config({"tags": ["my_chain"]})max_events = 0async for event in chain.astream_events('output a list of the countries france, spain and japan and their populations in JSON format. Use a dict with an outer key of "countries" which contains a list of countries. Each country should have the key `name` and `population`',version="v2",include_tags=["my_chain"],):print(event)max_events += 1if max_events > 10:# Truncate outputprint("...")break
{'event': 'on_chain_start', 'data': {'input': 'output a list of the countries france, spain and japan and their populations in JSON format. Use a dict with an outer key of "countries" which contains a list of countries. Each country should have the key `name` and `population`'}, 'name': 'RunnableSequence', 'tags': ['my_chain'], 'run_id': 'fd68dd64-7a4d-4bdb-a0c2-ee592db0d024', 'metadata': {}}{'event': 'on_chat_model_start', 'data': {'input': {'messages': [[HumanMessage(content='output a list of the countries france, spain and japan and their populations in JSON format. Use a dict with an outer key of "countries" which contains a list of countries. Each country should have the key `name` and `population`')]]}}, 'name': 'ChatAnthropic', 'tags': ['seq:step:1', 'my_chain'], 'run_id': 'efd3c8af-4be5-4f6c-9327-e3f9865dd1cd', 'metadata': {}}{'event': 'on_chat_model_stream', 'data': {'chunk': AIMessageChunk(content='{', id='run-efd3c8af-4be5-4f6c-9327-e3f9865dd1cd')}, 'run_id': 'efd3c8af-4be5-4f6c-9327-e3f9865dd1cd', 'name': 'ChatAnthropic', 'tags': ['seq:step:1', 'my_chain'], 'metadata': {}}{'event': 'on_parser_start', 'data': {}, 'name': 'JsonOutputParser', 'tags': ['seq:step:2', 'my_chain'], 'run_id': 'afde30b9-beac-4b36-b4c7-dbbe423ddcdb', 'metadata': {}}{'event': 'on_parser_stream', 'data': {'chunk': {}}, 'run_id': 'afde30b9-beac-4b36-b4c7-dbbe423ddcdb', 'name': 'JsonOutputParser', 'tags': ['seq:step:2', 'my_chain'], 'metadata': {}}{'event': 'on_chain_stream', 'data': {'chunk': {}}, 'run_id': 'fd68dd64-7a4d-4bdb-a0c2-ee592db0d024', 'name': 'RunnableSequence', 'tags': ['my_chain'], 'metadata': {}}{'event': 'on_chat_model_stream', 'data': {'chunk': AIMessageChunk(content='\n ', id='run-efd3c8af-4be5-4f6c-9327-e3f9865dd1cd')}, 'run_id': 'efd3c8af-4be5-4f6c-9327-e3f9865dd1cd', 'name': 'ChatAnthropic', 'tags': ['seq:step:1', 'my_chain'], 'metadata': {}}{'event': 'on_chat_model_stream', 'data': {'chunk': AIMessageChunk(content='"', id='run-efd3c8af-4be5-4f6c-9327-e3f9865dd1cd')}, 'run_id': 'efd3c8af-4be5-4f6c-9327-e3f9865dd1cd', 'name': 'ChatAnthropic', 'tags': ['seq:step:1', 'my_chain'], 'metadata': {}}{'event': 'on_chat_model_stream', 'data': {'chunk': AIMessageChunk(content='countries', id='run-efd3c8af-4be5-4f6c-9327-e3f9865dd1cd')}, 'run_id': 'efd3c8af-4be5-4f6c-9327-e3f9865dd1cd', 'name': 'ChatAnthropic', 'tags': ['seq:step:1', 'my_chain'], 'metadata': {}}{'event': 'on_chat_model_stream', 'data': {'chunk': AIMessageChunk(content='":', id='run-efd3c8af-4be5-4f6c-9327-e3f9865dd1cd')}, 'run_id': 'efd3c8af-4be5-4f6c-9327-e3f9865dd1cd', 'name': 'ChatAnthropic', 'tags': ['seq:step:1', 'my_chain'], 'metadata': {}}{'event': 'on_chat_model_stream', 'data': {'chunk': AIMessageChunk(content=' [', id='run-efd3c8af-4be5-4f6c-9327-e3f9865dd1cd')}, 'run_id': 'efd3c8af-4be5-4f6c-9327-e3f9865dd1cd', 'name': 'ChatAnthropic', 'tags': ['seq:step:1', 'my_chain'], 'metadata': {}}...
非流式组件
还记得一些组件因为不操作输入流而不能很好地进行流式传输吗?
尽管这样的组件在使用astream时可能会中断最终输出的流式传输,但astream_events仍然会从支持流式传输的中间步骤产生流事件!
# Function that does not support streaming.# It operates on the finalizes inputs rather than# operating on the input stream.def _extract_country_names(inputs):"""A function that does not operates on input streams and breaks streaming."""if not isinstance(inputs, dict):return ""if "countries" not in inputs:return ""countries = inputs["countries"]if not isinstance(countries, list):return ""country_names = [country.get("name") for country in countries if isinstance(country, dict)]return country_nameschain = (model | JsonOutputParser() | _extract_country_names) # This parser only works with OpenAI right now
正如预期的那样,astream API 没有正确工作,因为 _extract_country_names 不在流上操作。
async for chunk in chain.astream("output a list of the countries france, spain and japan and their populations in JSON format. "'Use a dict with an outer key of "countries" which contains a list of countries. '"Each country should have the key `name` and `population`",):print(chunk, flush=True)
['France', 'Spain', 'Japan']
好的,让我们确认一下,使用 astream_events 我们仍然可以从模型和解析器那里看到流式输出。
num_events = 0async for event in chain.astream_events("output a list of the countries france, spain and japan and their populations in JSON format. "'Use a dict with an outer key of "countries" which contains a list of countries. '"Each country should have the key `name` and `population`",version="v2",):kind = event["event"]if kind == "on_chat_model_stream":print(f"Chat model chunk: {repr(event['data']['chunk'].content)}",flush=True,)if kind == "on_parser_stream":print(f"Parser chunk: {event['data']['chunk']}", flush=True)num_events += 1if num_events > 30:# Truncate the outputprint("...")break
Chat model chunk: '{'Parser chunk: {}Chat model chunk: '\n 'Chat model chunk: '"'Chat model chunk: 'countries'Chat model chunk: '":'Chat model chunk: ' ['Parser chunk: {'countries': []}Chat model chunk: '\n 'Chat model chunk: '{'Parser chunk: {'countries': [{}]}Chat model chunk: '\n 'Chat model chunk: '"'Chat model chunk: 'name'Chat model chunk: '":'Chat model chunk: ' "'Parser chunk: {'countries': [{'name': ''}]}Chat model chunk: 'France'Parser chunk: {'countries': [{'name': 'France'}]}Chat model chunk: '",'Chat model chunk: '\n 'Chat model chunk: '"'Chat model chunk: 'population'Chat model chunk: '":'Chat model chunk: ' 'Chat model chunk: '67'Parser chunk: {'countries': [{'name': 'France', 'population': 67}]}...
传播回调
注意
如果您在工具中调用可运行的组件,您需要将回调传播到该组件;否则,将不会生成任何流事件。
注意
当使用 RunnableLambdas 或 @chain 装饰器时,回调会在后台自动传播。
from langchain_core.runnables import RunnableLambdafrom langchain_core.tools import tooldef reverse_word(word: str):return word[::-1]reverse_word = RunnableLambda(reverse_word)@tooldef bad_tool(word: str):"""Custom tool that doesn't propagate callbacks."""return reverse_word.invoke(word)async for event in bad_tool.astream_events("hello", version="v2"):print(event)
API 参考:RunnableLambda | tool
{'event': 'on_tool_start', 'data': {'input': 'hello'}, 'name': 'bad_tool', 'tags': [], 'run_id': 'ea900472-a8f7-425d-b627-facdef936ee8', 'metadata': {}}{'event': 'on_chain_start', 'data': {'input': 'hello'}, 'name': 'reverse_word', 'tags': [], 'run_id': '77b01284-0515-48f4-8d7c-eb27c1882f86', 'metadata': {}}{'event': 'on_chain_end', 'data': {'output': 'olleh', 'input': 'hello'}, 'run_id': '77b01284-0515-48f4-8d7c-eb27c1882f86', 'name': 'reverse_word', 'tags': [], 'metadata': {}}{'event': 'on_tool_end', 'data': {'output': 'olleh'}, 'run_id': 'ea900472-a8f7-425d-b627-facdef936ee8', 'name': 'bad_tool', 'tags': [], 'metadata': {}}
如果你在 Runnable Lambdas 或 @chain 中调用可运行的组件,那么回调将会自动代表你传递。
from langchain_core.runnables import RunnableLambdaasync def reverse_and_double(word: str):return await reverse_word.ainvoke(word) * 2reverse_and_double = RunnableLambda(reverse_and_double)await reverse_and_double.ainvoke("1234")async for event in reverse_and_double.astream_events("1234", version="v2"):print(event)
API 参考:RunnableLambda
{'event': 'on_chain_start', 'data': {'input': '1234'}, 'name': 'reverse_and_double', 'tags': [], 'run_id': '03b0e6a1-3e60-42fc-8373-1e7829198d80', 'metadata': {}}{'event': 'on_chain_start', 'data': {'input': '1234'}, 'name': 'reverse_word', 'tags': [], 'run_id': '5cf26fc8-840b-4642-98ed-623dda28707a', 'metadata': {}}{'event': 'on_chain_end', 'data': {'output': '4321', 'input': '1234'}, 'run_id': '5cf26fc8-840b-4642-98ed-623dda28707a', 'name': 'reverse_word', 'tags': [], 'metadata': {}}{'event': 'on_chain_stream', 'data': {'chunk': '43214321'}, 'run_id': '03b0e6a1-3e60-42fc-8373-1e7829198d80', 'name': 'reverse_and_double', 'tags': [], 'metadata': {}}{'event': 'on_chain_end', 'data': {'output': '43214321'}, 'run_id': '03b0e6a1-3e60-42fc-8373-1e7829198d80', 'name': 'reverse_and_double', 'tags': [], 'metadata': {}}
使用 @chain 装饰器:
from langchain_core.runnables import chain@chainasync def reverse_and_double(word: str):return await reverse_word.ainvoke(word) * 2await reverse_and_double.ainvoke("1234")async for event in reverse_and_double.astream_events("1234", version="v2"):print(event)
API 参考:chain
{'event': 'on_chain_start', 'data': {'input': '1234'}, 'name': 'reverse_and_double', 'tags': [], 'run_id': '1bfcaedc-f4aa-4d8e-beee-9bba6ef17008', 'metadata': {}}{'event': 'on_chain_start', 'data': {'input': '1234'}, 'name': 'reverse_word', 'tags': [], 'run_id': '64fc99f0-5d7d-442b-b4f5-4537129f67d1', 'metadata': {}}{'event': 'on_chain_end', 'data': {'output': '4321', 'input': '1234'}, 'run_id': '64fc99f0-5d7d-442b-b4f5-4537129f67d1', 'name': 'reverse_word', 'tags': [], 'metadata': {}}{'event': 'on_chain_stream', 'data': {'chunk': '43214321'}, 'run_id': '1bfcaedc-f4aa-4d8e-beee-9bba6ef17008', 'name': 'reverse_and_double', 'tags': [], 'metadata': {}}{'event': 'on_chain_end', 'data': {'output': '43214321'}, 'run_id': '1bfcaedc-f4aa-4d8e-beee-9bba6ef17008', 'name': 'reverse_and_double', 'tags': [], 'metadata': {}}
后续步骤
现在你已经学会了使用LangChain流式传输最终输出和内部步骤的一些方法。
要了解更多,请查看本节中的其他操作指南,或查看Langchain表达式语言的概念指南。
