개발자로서 현장에서 일하면서 새로 접하는 기술들이나 알게된 정보 등을 정리하기 위한 블로그입니다. 운 좋게 미국에서 큰 회사들의 프로젝트에서 컬설턴트로 일하고 있어서 새로운 기술들을 접할 기회가 많이 있습니다. 미국의 IT 프로젝트에서 사용되는 툴들에 대해 많은 분들과 정보를 공유하고 싶습니다.
To best understand the agent framework, let’s build an agent that has two tools: one to look things up online, and one to look up specific data that we’ve loaded into a index.
에이전트 프레임워크를 가장 잘 이해하기 위해 두 가지 도구가 있는 에이전트를 구축해 보겠습니다. 하나는 온라인으로 항목을 조회하는 도구이고 다른 하나는 인덱스에 로드한 특정 데이터를 조회하는 도구입니다.
This will assume knowledge ofLLMsandretrievalso if you haven’t already explored those sections, it is recommended you do so.
이는 LLM 및 검색에 대한 지식을 가정하므로 해당 섹션을 아직 탐색하지 않았다면 탐색하는 것이 좋습니다.
By definition, agents take a self-determined, input-dependent sequence of steps before returning a user-facing output. This makes debugging these systems particularly tricky, and observability particularly important.LangSmithis especially useful for such cases.
정의에 따르면 에이전트는 사용자에게 표시되는 출력을 반환하기 전에 입력에 따라 자체 결정된 일련의 단계를 수행합니다. 이로 인해 이러한 시스템을 디버깅하는 것이 특히 까다로워지고 관찰 가능성이 특히 중요해집니다. LangSmith는 이러한 경우에 특히 유용합니다.
When building with LangChain, all steps will automatically be traced in LangSmith. To set up LangSmith we just need set the following environment variables:
LangChain으로 구축할 때 모든 단계는 LangSmith에서 자동으로 추적됩니다. LangSmith를 설정하려면 다음 환경 변수만 설정하면 됩니다.
We first need to create the tools we want to use. We will use two tools:Tavily(to search online) and then a retriever over a local index we will create
먼저 사용하려는 도구를 만들어야 합니다. 우리는 두 가지 도구를 사용할 것입니다: Tavily(온라인 검색용)와 우리가 생성할 로컬 인덱스에 대한 검색기
We have a built-in tool in LangChain to easily use Tavily search engine as tool. Note that this requires an API key - they have a free tier, but if you don’t have one or don’t want to create one, you can always ignore this step.
우리는 Tavily 검색 엔진을 도구로 쉽게 사용할 수 있도록 LangChain에 내장된 도구를 가지고 있습니다. 이를 위해서는 API 키가 필요합니다. 무료 등급이 있지만 API 키가 없거나 생성하고 싶지 않은 경우 언제든지 이 단계를 무시할 수 있습니다.
Once you create your API key, you will need to export that as:
API 키를 생성한 후에는 이를 다음과 같이 내보내야 합니다.
export TAVILY_API_KEY="..."
from langchain_community.tools.tavily_search import TavilySearchResults
search = TavilySearchResults()
search.invoke("what is the weather in SF")
[{'url': 'https://www.metoffice.gov.uk/weather/forecast/9q8yym8kr',
'content': 'Thu 11 Jan Thu 11 Jan Seven day forecast for San Francisco San Francisco (United States of America) weather Find a forecast Sat 6 Jan Sat 6 Jan Sun 7 Jan Sun 7 Jan Mon 8 Jan Mon 8 Jan Tue 9 Jan Tue 9 Jan Wed 10 Jan Wed 10 Jan Thu 11 Jan Find a forecast Please choose your location from the nearest places to : Forecast days Today Today Sat 6 Jan Sat 6 JanSan Francisco 7 day weather forecast including weather warnings, temperature, rain, wind, visibility, humidity and UV ... (11 January 2024) Time 00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:00 09:00 10:00 11:00 12:00 ... Oakland Int. 11.5 miles; San Francisco International 11.5 miles; Corte Madera 12.3 miles; Redwood City 23.4 miles;'},
{'url': 'https://www.latimes.com/travel/story/2024-01-11/east-brother-light-station-lighthouse-california',
'content': "May 18, 2023 Jan. 4, 2024 Subscribe for unlimited accessSite Map Follow Us MORE FROM THE L.A. TIMES Jan. 8, 2024 Travel & Experiences This must be Elysian Valley (a.k.a. Frogtown) Jan. 5, 2024 Food June 30, 2023The East Brother Light Station in the San Francisco Bay is not a destination for everyone. ... Jan. 11, 2024 3 AM PT ... Champagne and hors d'oeuvres are served in late afternoon — outdoors if ..."}]
We will also create a retriever over some data of our own. For a deeper explanation of each step here, seethis section.
우리는 또한 우리 자신의 일부 데이터에 대한 검색기를 만들 것입니다. 각 단계에 대한 자세한 설명은 이 섹션을 참조하세요.
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain_community.document_loaders import WebBaseLoader
from langchain_community.vectorstores import FAISS
from langchain_openai import OpenAIEmbeddings
loader = WebBaseLoader("https://docs.smith.langchain.com/overview")
docs = loader.load()
documents = RecursiveCharacterTextSplitter(
chunk_size=1000, chunk_overlap=200
).split_documents(docs)
vector = FAISS.from_documents(documents, OpenAIEmbeddings())
retriever = vector.as_retriever()
retriever.get_relevant_documents("how to upload a dataset")[0]
Document(page_content="dataset uploading.Once we have a dataset, how can we use it to test changes to a prompt or chain? The most basic approach is to run the chain over the data points and visualize the outputs. Despite technological advancements, there still is no substitute for looking at outputs by eye. Currently, running the chain over the data points needs to be done client-side. The LangSmith client makes it easy to pull down a dataset and then run a chain over them, logging the results to a new project associated with the dataset. From there, you can review them. We've made it easy to assign feedback to runs and mark them as correct or incorrect directly in the web app, displaying aggregate statistics for each test project.We also make it easier to evaluate these runs. To that end, we've added a set of evaluators to the open-source LangChain library. These evaluators can be specified when initiating a test run and will evaluate the results once the test run completes. If we’re being honest, most of", metadata={'source': 'https://docs.smith.langchain.com/overview', 'title': 'LangSmith Overview and User Guide | 🦜️🛠️ LangSmith', 'description': 'Building reliable LLM applications can be challenging. LangChain simplifies the initial setup, but there is still work needed to bring the performance of prompts, chains and agents up the level where they are reliable enough to be used in production.', 'language': 'en'})
Now that we have populated our index that we will do doing retrieval over, we can easily turn it into a tool (the format needed for an agent to properly use it)
이제 검색을 수행할 인덱스를 채웠으므로 이를 도구(에이전트가 올바르게 사용하는 데 필요한 형식)로 쉽게 전환할 수 있습니다.
from langchain.tools.retriever import create_retriever_tool
retriever_tool = create_retriever_tool(
retriever,
"langsmith_search",
"Search for information about LangSmith. For any questions about LangSmith, you must use this tool!",
)
Now that we have defined the tools, we can create the agent. We will be using an OpenAI Functions agent - for more information on this type of agent, as well as other options, seethis guide
from langchain import hub
# Get the prompt to use - you can modify this!
prompt = hub.pull("hwchase17/openai-functions-agent")
prompt.messages
[SystemMessagePromptTemplate(prompt=PromptTemplate(input_variables=[], template='You are a helpful assistant')),
MessagesPlaceholder(variable_name='chat_history', optional=True),
HumanMessagePromptTemplate(prompt=PromptTemplate(input_variables=['input'], template='{input}')),
MessagesPlaceholder(variable_name='agent_scratchpad')]
Now, we can initalize the agent with the LLM, the prompt, and the tools. The agent is responsible for taking in input and deciding what actions to take. Crucially, the Agent does not execute those actions - that is done by the AgentExecutor (next step). For more information about how to think about these components, see ourconceptual guide
이제 LLM, 프롬프트 및 도구를 사용하여 에이전트를 초기화할 수 있습니다. 에이전트는 입력을 받아들이고 어떤 조치를 취할지 결정하는 역할을 담당합니다. 결정적으로 에이전트는 이러한 작업을 실행하지 않습니다. 이는 AgentExecutor(다음 단계)에 의해 수행됩니다. 이러한 구성 요소에 대해 생각하는 방법에 대한 자세한 내용은 개념 가이드를 참조하세요.
from langchain.agents import create_openai_functions_agent
agent = create_openai_functions_agent(llm, tools, prompt)
Finally, we combine the agent (the brains) with the tools inside the AgentExecutor (which will repeatedly call the agent and execute tools). For more information about how to think about these components, see ourconceptual guide
We can now run the agent on a few queries! Note that for now, these are allstatelessqueries (it won’t remember previous interactions).
이제 몇 가지 쿼리에 대해 에이전트를 실행할 수 있습니다! 현재로서는 이는 모두 상태 비저장 쿼리입니다(이전 상호 작용을 기억하지 않음).
agent_executor.invoke({"input": "hi!"})
> Entering new AgentExecutor chain...
Hello! How can I assist you today?
> Finished chain.
{'input': 'hi!', 'output': 'Hello! How can I assist you today?'}
agent_executor.invoke({"input": "how can langsmith help with testing?"})
> Entering new AgentExecutor chain...
Invoking: `langsmith_search` with `{'query': 'LangSmith testing'}`
[Document(page_content='LangSmith Overview and User Guide | 🦜️🛠️ LangSmith', metadata={'source': 'https://docs.smith.langchain.com/overview', 'title': 'LangSmith Overview and User Guide | 🦜️🛠️ LangSmith', 'description': 'Building reliable LLM applications can be challenging. LangChain simplifies the initial setup, but there is still work needed to bring the performance of prompts, chains and agents up the level where they are reliable enough to be used in production.', 'language': 'en'}), Document(page_content='Skip to main content🦜️🛠️ LangSmith DocsPython DocsJS/TS DocsSearchGo to AppLangSmithOverviewTracingTesting & EvaluationOrganizationsHubLangSmith CookbookOverviewOn this pageLangSmith Overview and User GuideBuilding reliable LLM applications can be challenging. LangChain simplifies the initial setup, but there is still work needed to bring the performance of prompts, chains and agents up the level where they are reliable enough to be used in production.Over the past two months, we at LangChain have been building and using LangSmith with the goal of bridging this gap. This is our tactical user guide to outline effective ways to use LangSmith and maximize its benefits.On by default\u200bAt LangChain, all of us have LangSmith’s tracing running in the background by default. On the Python side, this is achieved by setting environment variables, which we establish whenever we launch a virtual environment or open our bash shell and leave them set. The same principle applies to most JavaScript', metadata={'source': 'https://docs.smith.langchain.com/overview', 'title': 'LangSmith Overview and User Guide | 🦜️🛠️ LangSmith', 'description': 'Building reliable LLM applications can be challenging. LangChain simplifies the initial setup, but there is still work needed to bring the performance of prompts, chains and agents up the level where they are reliable enough to be used in production.', 'language': 'en'}), Document(page_content='You can also quickly edit examples and add them to datasets to expand the surface area of your evaluation sets or to fine-tune a model for improved quality or reduced costs.Monitoring\u200bAfter all this, your app might finally ready to go in production. LangSmith can also be used to monitor your application in much the same way that you used for debugging. You can log all traces, visualize latency and token usage statistics, and troubleshoot specific issues as they arise. Each run can also be assigned string tags or key-value metadata, allowing you to attach correlation ids or AB test variants, and filter runs accordingly.We’ve also made it possible to associate feedback programmatically with runs. This means that if your application has a thumbs up/down button on it, you can use that to log feedback back to LangSmith. This can be used to track performance over time and pinpoint under performing data points, which you can subsequently add to a dataset for future testing — mirroring the', metadata={'source': 'https://docs.smith.langchain.com/overview', 'title': 'LangSmith Overview and User Guide | 🦜️🛠️ LangSmith', 'description': 'Building reliable LLM applications can be challenging. LangChain simplifies the initial setup, but there is still work needed to bring the performance of prompts, chains and agents up the level where they are reliable enough to be used in production.', 'language': 'en'}), Document(page_content='inputs, and see what happens. At some point though, our application is performing\nwell and we want to be more rigorous about testing changes. We can use a dataset\nthat we’ve constructed along the way (see above). Alternatively, we could spend some\ntime constructing a small dataset by hand. For these situations, LangSmith simplifies', metadata={'source': 'https://docs.smith.langchain.com/overview', 'title': 'LangSmith Overview and User Guide | 🦜️🛠️ LangSmith', 'description': 'Building reliable LLM applications can be challenging. LangChain simplifies the initial setup, but there is still work needed to bring the performance of prompts, chains and agents up the level where they are reliable enough to be used in production.', 'language': 'en'})]LangSmith can help with testing in several ways. Here are some ways LangSmith can assist with testing:
1. Tracing: LangSmith provides tracing capabilities that can be used to monitor and debug your application during testing. You can log all traces, visualize latency and token usage statistics, and troubleshoot specific issues as they arise.
2. Evaluation: LangSmith allows you to quickly edit examples and add them to datasets to expand the surface area of your evaluation sets. This can help you test and fine-tune your models for improved quality or reduced costs.
3. Monitoring: Once your application is ready for production, LangSmith can be used to monitor your application. You can log feedback programmatically with runs, track performance over time, and pinpoint underperforming data points. This information can be used to improve your application and add to datasets for future testing.
4. Rigorous Testing: When your application is performing well and you want to be more rigorous about testing changes, LangSmith can simplify the process. You can use existing datasets or construct small datasets by hand to test different scenarios and evaluate the performance of your application.
For more detailed information on how to use LangSmith for testing, you can refer to the [LangSmith Overview and User Guide](https://docs.smith.langchain.com/overview).
> Finished chain.
{'input': 'how can langsmith help with testing?',
'output': 'LangSmith can help with testing in several ways. Here are some ways LangSmith can assist with testing:\n\n1. Tracing: LangSmith provides tracing capabilities that can be used to monitor and debug your application during testing. You can log all traces, visualize latency and token usage statistics, and troubleshoot specific issues as they arise.\n\n2. Evaluation: LangSmith allows you to quickly edit examples and add them to datasets to expand the surface area of your evaluation sets. This can help you test and fine-tune your models for improved quality or reduced costs.\n\n3. Monitoring: Once your application is ready for production, LangSmith can be used to monitor your application. You can log feedback programmatically with runs, track performance over time, and pinpoint underperforming data points. This information can be used to improve your application and add to datasets for future testing.\n\n4. Rigorous Testing: When your application is performing well and you want to be more rigorous about testing changes, LangSmith can simplify the process. You can use existing datasets or construct small datasets by hand to test different scenarios and evaluate the performance of your application.\n\nFor more detailed information on how to use LangSmith for testing, you can refer to the [LangSmith Overview and User Guide](https://docs.smith.langchain.com/overview).'}
As mentioned earlier, this agent is stateless. This means it does not remember previous interactions. To give it memory we need to pass in previouschat_history. Note: it needs to be calledchat_historybecause of the prompt we are using. If we use a different prompt, we could change the variable name
앞서 언급했듯이 이 에이전트는 Stateless입니다. 이는 이전 상호작용을 기억하지 못한다는 것을 의미합니다. 메모리를 제공하려면 이전 chat_history를 전달해야 합니다. 참고: 우리가 사용하는 프롬프트 때문에 chat_history라고 불러야 합니다. 다른 프롬프트를 사용하면 변수 이름을 변경할 수 있습니다.
# Here we pass in an empty list of messages for chat_history because it is the first message in the chat
agent_executor.invoke({"input": "hi! my name is bob", "chat_history": []})
> Entering new AgentExecutor chain...
Hello Bob! How can I assist you today?
> Finished chain.
{'input': 'hi! my name is bob',
'chat_history': [],
'output': 'Hello Bob! How can I assist you today?'}
from langchain_core.messages import AIMessage, HumanMessage
agent_executor.invoke(
{
"chat_history": [
HumanMessage(content="hi! my name is bob"),
AIMessage(content="Hello Bob! How can I assist you today?"),
],
"input": "what's my name?",
}
)
> Entering new AgentExecutor chain...
Your name is Bob.
> Finished chain.
{'chat_history': [HumanMessage(content='hi! my name is bob'),
AIMessage(content='Hello Bob! How can I assist you today?')],
'input': "what's my name?",
'output': 'Your name is Bob.'}
If we want to keep track of these messages automatically, we can wrap this in a RunnableWithMessageHistory. For more information on how to use this, seethis guide
이러한 메시지를 자동으로 추적하려면 이를 RunnableWithMessageHistory로 래핑할 수 있습니다. 이를 사용하는 방법에 대한 자세한 내용은 이 가이드를 참조하세요.
from langchain_community.chat_message_histories import ChatMessageHistory
from langchain_core.runnables.history import RunnableWithMessageHistory
message_history = ChatMessageHistory()
agent_with_chat_history = RunnableWithMessageHistory(
agent_executor,
# This is needed because in most real world scenarios, a session id is needed
# It isn't really used here because we are using a simple in memory ChatMessageHistory
lambda session_id: message_history,
input_messages_key="input",
history_messages_key="chat_history",
)
agent_with_chat_history.invoke(
{"input": "hi! I'm bob"},
# This is needed because in most real world scenarios, a session id is needed
# It isn't really used here because we are using a simple in memory ChatMessageHistory
config={"configurable": {"session_id": "<foo>"}},
)
> Entering new AgentExecutor chain...
Hello Bob! How can I assist you today?
> Finished chain.
{'input': "hi! I'm bob",
'chat_history': [],
'output': 'Hello Bob! How can I assist you today?'}
agent_with_chat_history.invoke(
{"input": "what's my name?"},
# This is needed because in most real world scenarios, a session id is needed
# It isn't really used here because we are using a simple in memory ChatMessageHistory
config={"configurable": {"session_id": "<foo>"}},
)
> Entering new AgentExecutor chain...
Your name is Bob.
> Finished chain.
{'input': "what's my name?",
'chat_history': [HumanMessage(content="hi! I'm bob"),
AIMessage(content='Hello Bob! How can I assist you today?')],
'output': 'Your name is Bob.'}
That’s a wrap! In this quick start we covered how to create a simple agent. Agents are a complex topic, and there’s lot to learn! Head back to themain agent pageto find more resources on conceptual guides, different types of agents, how to create custom tools, and more!
마무리입니다! 이 빠른 시작에서는 간단한 에이전트를 만드는 방법을 다루었습니다. 에이전트는 복잡한 주제이며 배울 것이 많습니다! 기본 에이전트 페이지로 돌아가서 개념 가이드, 다양한 유형의 에이전트, 사용자 정의 도구를 만드는 방법 등에 대한 추가 리소스를 찾아보세요!
