■ WebBaseLoader 클래스의 생성자에서 web_paths/bs_kwargs 인자를 사용해 웹 문서 로더를 만드는 방법을 보여준다. ▶ main.py
|
1 2 3 4 5 6 7 8 9 10 |
import bs4 from langchain_community.document_loaders import WebBaseLoader webBaseLoader = WebBaseLoader( web_paths = ("https://lilianweng.github.io/posts/2023-06-23-agent/",), bs_kwargs = dict(parse_only = bs4.SoupStrainer(class_ = ("post-content", "post-title", "post-header"))) ) |
■ RunnableSequence 클래스를 사용해 문서 검색 모듈과 질문/답변 생성 모듈을 연결하는 방법을 보여준다. ※ [create_retrieval_chain 함수 : 문서 검색 모듈과 질문/답변 생성
■ PDFMinerPDFasHTMLLoader 클래스의 load 메소드를 사용해 PDF 파일 문서를 로드하는 방법을 보여준다. ▶ main.py
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 |
import re from bs4 import BeautifulSoup from langchain_community.document_loaders import PDFMinerPDFasHTMLLoader from langchain_core.documents import Document pdfMinerPDFasHTMLLoader = PDFMinerPDFasHTMLLoader("nke-10k-2023.pdf") documentList = pdfMinerPDFasHTMLLoader.load() document = documentList[0] beautifulSoup = BeautifulSoup(document.page_content, "html.parser") resultSet = beautifulSoup.find_all("div") currentFontSize = None currentText = "" snippetList = [] # (text, fontSize) 튜플 리스트 for result in resultSet: span = result.find("span") if not span: continue style = span.get("style") if not style: continue fontSizeString = re.findall("font-size:(\d+)px", style) if not fontSizeString: continue fontSize = int(fontSizeString[0]) if not currentFontSize: currentFontSize = fontSize if fontSize == currentFontSize: currentText += result.text else: snippetList.append((currentText, currentFontSize)) currentFontSize = fontSize currentText = result.text snippetList.append((currentText, currentFontSize)) # for snippet in snippetList[:3]: # print(snippet) """ ('FORM 10-K\nFORM 10-K\n', 29) ('UNITED STATES\nSECURITIES AND EXCHANGE COMMISSION\nWashington, D.C. 20549\n', 7) ('FORM 10-K \n', 10) """ currentIndex = -1 semanticSnippetDocumentList = [] # 가정 : 제목은 해당 내용보다 글꼴 크기가 더 크다. for snippet in snippetList: # 현재 조각의 글꼴 크기 > 이전 섹션의 제목 => 새 제목인 경우 if (not semanticSnippetDocumentList or snippet[1] > semanticSnippetDocumentList[currentIndex].metadata["heading_font"]): metadataDictionary = {"heading" : snippet[0], "content_font" : 0, "heading_font" : snippet[1]} metadataDictionary.update(document.metadata) semanticSnippetDocumentList.append(Document(page_content = "", metadata = metadataDictionary)) currentIndex += 1 continue # 현재 조각의 글꼴 크기 <= 이전 섹션의 콘텐츠 => 콘텐츠가 동일한 섹션에 속하는 경우 # (필요한 경우 하위 섹션에 대해 트리와 같은 구조를 만들 수도 있지만 이는 더 많은 생각이 필요할 수 있으며 데이터에 따라 다를 수 있다) if (not semanticSnippetDocumentList[currentIndex].metadata["content_font"] or snippet[1] <= semanticSnippetDocumentList[currentIndex].metadata["content_font"]): semanticSnippetDocumentList[currentIndex].page_content += snippet[0] semanticSnippetDocumentList[currentIndex].metadata["content_font"] = max(snippet[1], semanticSnippetDocumentList[currentIndex].metadata["content_font"]) continue # 현재 조각의 글꼴 크기가 이전 섹션의 내용보다 크지만 이전 섹션의 제목보다 작은 경우 새 섹션을 만드는 것 # (예 : PDF 제목의 글꼴 크기가 가장 높지만 모든 섹션을 포함하는 것을 원하지 않는다) metadataDictionary = {"heading" : snippet[0], "content_font" : 0, "heading_font" : snippet[1]} metadataDictionary.update(document.metadata) semanticSnippetDocumentList.append(Document(page_content = "", metadata = metadataDictionary)) currentIndex += 1 for semanticSnippetDocument in semanticSnippetDocumentList[:3]: print(semanticSnippetDocument.page_content) print() """ UNITED STATES SECURITIES AND EXCHANGE COMMISSION Washington, D.C. 20549 (Mark One) ☑ ANNUAL REPORT PURSUANT TO SECTION 13 OR 15(D) OF THE SECURITIES EXCHANGE ACT OF 1934 FOR THE FISCAL YEAR ENDED MAY 31, 2023 OR ☐ TRANSITION REPORT PURSUANT TO SECTION 13 OR 15(D) OF THE SECURITIES EXCHANGE ACT OF 1934 FOR THE TRANSITION PERIOD FROM TO . Commission File No. 1-10635 (Exact name of Registrant as specified in its charter) Oregon (State or other jurisdiction of incorporation) 93-0584541 (IRS Employer Identification No.) One Bowerman Drive, Beaverton, Oregon 97005-6453 (Address of principal executive offices and zip code) (503) 671-6453 (Registrant's telephone number, including area code) SECURITIES REGISTERED PURSUANT TO SECTION 12(B) OF THE ACT: Class B Common Stock (Title of each class) NKE New York Stock Exchange (Trading symbol) (Name of each exchange on which registered) SECURITIES REGISTERED PURSUANT TO SECTION 12(G) OF THE ACT: NONE Indicate by check mark: • if the registrant is a well-known seasoned issuer, as defined in Rule 405 of the Securities Act. • if the registrant is not required to file reports pursuant to Section 13 or Section 15(d) of the Act. • whether the registrant (1) has filed all reports required to be filed by Section 13 or 15(d) of the Securities Exchange Act of 1934 during the preceding 12 months (or for such shorter period that the registrant was required to file such reports), and (2) has been subject to such filing requirements for the past 90 days. """ |
▶ requirements.txt
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 |
aiohttp==3.9.5 aiosignal==1.3.1 annotated-types==0.7.0 async-timeout==4.0.3 attrs==23.2.0 beautifulsoup4==4.12.3 bs4==0.0.2 certifi==2024.6.2 cffi==1.16.0 charset-normalizer==3.3.2 cryptography==42.0.8 dataclasses-json==0.6.7 frozenlist==1.4.1 greenlet==3.0.3 idna==3.7 jsonpatch==1.33 jsonpointer==3.0.0 langchain==0.2.6 langchain-community==0.2.6 langchain-core==0.2.10 langchain-text-splitters==0.2.2 langsmith==0.1.82 marshmallow==3.21.3 multidict==6.0.5 mypy-extensions==1.0.0 numpy==1.26.4 orjson==3.10.5 packaging==24.1 pdfminer.six==20231228 pycparser==2.22 pydantic==2.7.4 pydantic_core==2.18.4 PyYAML==6.0.1 requests==2.32.3 soupsieve==2.5 SQLAlchemy==2.0.31 tenacity==8.4.2 typing-inspect==0.9.0 typing_extensions==4.12.2 urllib3==2.2.2 yarl==1.9.4 |
※ pip install langchain-community
■ PDFMinerPDFasHTMLLoader 클래스를 사용해 PDF 파일 로드 후 HTML 텍스트를 만드는 방법을 보여준다. ▶ main.py
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 |
from bs4 import BeautifulSoup from langchain_community.document_loaders import PDFMinerPDFasHTMLLoader pdfMinerPDFasHTMLLoader = PDFMinerPDFasHTMLLoader("nke-10k-2023.pdf") documentList = pdfMinerPDFasHTMLLoader.load() document = documentList[0] soup = BeautifulSoup(document.page_content, "html.parser") resultSet = soup.find_all("div") for result in resultSet[:2]: print(result) print() """ <div style="position:absolute; top:50px;"><a name="1">Page 1</a></div> <div style="position:absolute; border: textbox 1px solid; writing-mode:lr-tb; left:422px; top:192px; width:138px; height:29px;"><span style="font-family: FuturaStd-CondensedExtraBd; font-size:29px">FORM 10-K <br/>FORM 10-K <br/></span></div> """ |
▶ requirements.txt
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 |
aiohttp==3.9.5 aiosignal==1.3.1 annotated-types==0.7.0 async-timeout==4.0.3 attrs==23.2.0 beautifulsoup4==4.12.3 bs4==0.0.2 certifi==2024.6.2 cffi==1.16.0 charset-normalizer==3.3.2 cryptography==42.0.8 dataclasses-json==0.6.7 frozenlist==1.4.1 greenlet==3.0.3 idna==3.7 jsonpatch==1.33 jsonpointer==3.0.0 langchain==0.2.6 langchain-community==0.2.6 langchain-core==0.2.10 langchain-text-splitters==0.2.2 langsmith==0.1.82 marshmallow==3.21.3 multidict==6.0.5 mypy-extensions==1.0.0 numpy==1.26.4 orjson==3.10.5 packaging==24.1 pdfminer.six==20231228 pycparser==2.22 pydantic==2.7.4 pydantic_core==2.18.4 PyYAML==6.0.1 requests==2.32.3 soupsieve==2.5 SQLAlchemy==2.0.31 tenacity==8.4.2 typing-inspect==0.9.0 typing_extensions==4.12.2 urllib3==2.2.2 yarl==1.9.4 |
※ pip install
■ RunnableWithMessageHistory 클래스를 사용해 채팅 메시지 히스토리 내역을 출력하는 방법을 보여준다. ▶ main.py
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 |
import os import bs4 from langchain_openai import ChatOpenAI from langchain_core.prompts import ChatPromptTemplate from langchain.chains.combine_documents import create_stuff_documents_chain from langchain_community.document_loaders import WebBaseLoader from langchain_text_splitters import RecursiveCharacterTextSplitter from langchain_chroma import Chroma from langchain_openai import OpenAIEmbeddings from langchain.chains import create_retrieval_chain from langchain_core.prompts import MessagesPlaceholder from langchain.chains import create_history_aware_retriever from langchain_core.chat_history import BaseChatMessageHistory from langchain_community.chat_message_histories import ChatMessageHistory from langchain_core.runnables.history import RunnableWithMessageHistory from langchain_core.messages import AIMessage # OPANAI API 키를 설정한다. os.environ["OPENAI_API_KEY"] = "<OPENAI_API_KEY>" # OPENAI 챗봇 모델을 설정한다. chatOpenAI = ChatOpenAI(model = "gpt-3.5-turbo-0125") # 질문-응답 채팅 프롬프트 템플리트를 설정한다. qnaChatPromptTemplate = ChatPromptTemplate.from_messages( [ ("system", "You are an assistant for question-answering tasks. Use the following pieces of retrieved context to answer the question. If you don't know the answer, say that you don't know. Use three sentences maximum and keep the answer concise.\n\n{context}"), MessagesPlaceholder("chatHistoryList"), ("human", "{input}"), ] ) # 질문-응답 체인을 생성한다. qnaRunnableBinding = create_stuff_documents_chain(chatOpenAI, qnaChatPromptTemplate) # 웹 문서를 로딩한다. webBaseLoader = WebBaseLoader( web_paths = ("https://lilianweng.github.io/posts/2023-06-23-agent/",), bs_kwargs = dict(parse_only = bs4.SoupStrainer(class_ = ("post-content", "post-title", "post-header"))) ) documentList = webBaseLoader.load() # 문서를 분할한다. recursiveCharacterTextSplitter = RecursiveCharacterTextSplitter(chunk_size = 1000, chunk_overlap = 200) splitDocumentList = recursiveCharacterTextSplitter.split_documents(documentList) # 문서 임베딩 및 검색기를 설정한다. chroma = Chroma.from_documents(documents = splitDocumentList, embedding = OpenAIEmbeddings()) vectorStoreRetriever = chroma.as_retriever() # 질문 재구성 채팅 프롬프트 템플리트를 설정한다. contextualizeQuestionChatPromptTemplate = ChatPromptTemplate.from_messages( [ ("system", "Given a chat history and the latest user question which might reference context in the chat history, formulate a standalone question which can be understood without the chat history. Do NOT answer the question, just reformulate it if needed and otherwise return it as is."), MessagesPlaceholder("chatHistoryList"), ("human", "{input}"), ] ) # 히스토리 기반 검색기 체인을 생성한다. historyAwareRetrieverRunnableBinding = create_history_aware_retriever(chatOpenAI, vectorStoreRetriever, contextualizeQuestionChatPromptTemplate) # RAG 체인을 생성한다. ragRunnableBinding = create_retrieval_chain(historyAwareRetrieverRunnableBinding, qnaRunnableBinding) # 채팅 메시지 히스토리 딕셔너리를 설정한다. chatMessageHistoryDictionary = {} def getSessionHistory(sessionID : str) -> BaseChatMessageHistory: if sessionID not in chatMessageHistoryDictionary: chatMessageHistoryDictionary[sessionID] = ChatMessageHistory() return chatMessageHistoryDictionary[sessionID] # 대화형 RAG 실행 가능 객체를 설정한다. conversationalRAGRunnableBinding = RunnableWithMessageHistory( ragRunnableBinding, getSessionHistory, input_messages_key = "input", history_messages_key = "chatHistoryList", output_messages_key = "answer", ) # 질문에 대한 응답 수행을 수행한다. aiMessage1 = conversationalRAGRunnableBinding.invoke( {"input" : "What is Task Decomposition?"}, config = {"configurable" : {"session_id" : "session1"}} ) aiMessage2 = conversationalRAGRunnableBinding.invoke( {"input" : "What are common ways of doing it?"}, config = {"configurable" : {"session_id" : "session1"}}, ) # 채팅 메시지 히스토리 내역을 출력한다. for message in chatMessageHistoryDictionary["session1"].messages: if isinstance(message, AIMessage): prefix = "AI" else: prefix = "User" print(f"{prefix} : {message.content}\n") """ User : What is Task Decomposition? AI : Task decomposition is a technique used to break down complex tasks into smaller, more manageable steps. This process helps agents or models tackle difficult tasks by dividing them into simpler subtasks, making it easier to plan and execute the overall task effectively. Task decomposition can be achieved through methods like prompting the model with step-by-step instructions, task-specific guidance, or human input. User : What are common ways of doing it? AI : Task decomposition can be achieved through various methods, including prompting the model with step-by-step instructions such as "Steps for XYZ" or "What are the subgoals for achieving XYZ?", providing task-specific guidance like "Write a story outline" for writing a novel, or incorporating human inputs to guide the decomposition process. These approaches help break down complex tasks into smaller, more manageable steps, enabling agents or models to effectively plan and execute the overall task. """ |
▶ requirements.txt
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 |
aiohttp==3.9.5 aiosignal==1.3.1 annotated-types==0.7.0 anyio==4.4.0 asgiref==3.8.1 async-timeout==4.0.3 attrs==23.2.0 backoff==2.2.1 bcrypt==4.1.3 beautifulsoup4==4.12.3 bs4==0.0.2 build==1.2.1 cachetools==5.3.3 certifi==2024.6.2 charset-normalizer==3.3.2 chroma-hnswlib==0.7.3 chromadb==0.5.0 click==8.1.7 coloredlogs==15.0.1 dataclasses-json==0.6.7 Deprecated==1.2.14 distro==1.9.0 dnspython==2.6.1 email_validator==2.1.1 exceptiongroup==1.2.1 fastapi==0.111.0 fastapi-cli==0.0.4 filelock==3.14.0 flatbuffers==24.3.25 frozenlist==1.4.1 fsspec==2024.6.0 google-auth==2.30.0 googleapis-common-protos==1.63.1 greenlet==3.0.3 grpcio==1.64.1 h11==0.14.0 httpcore==1.0.5 httptools==0.6.1 httpx==0.27.0 huggingface-hub==0.23.3 humanfriendly==10.0 idna==3.7 importlib_metadata==7.1.0 importlib_resources==6.4.0 Jinja2==3.1.4 jsonpatch==1.33 jsonpointer==3.0.0 kubernetes==30.1.0 langchain==0.2.3 langchain-chroma==0.1.1 langchain-community==0.2.4 langchain-core==0.2.5 langchain-openai==0.1.8 langchain-text-splitters==0.2.1 langsmith==0.1.77 markdown-it-py==3.0.0 MarkupSafe==2.1.5 marshmallow==3.21.3 mdurl==0.1.2 mmh3==4.1.0 monotonic==1.6 mpmath==1.3.0 multidict==6.0.5 mypy-extensions==1.0.0 numpy==1.26.4 oauthlib==3.2.2 onnxruntime==1.18.0 openai==1.33.0 opentelemetry-api==1.25.0 opentelemetry-exporter-otlp-proto-common==1.25.0 opentelemetry-exporter-otlp-proto-grpc==1.25.0 opentelemetry-instrumentation==0.46b0 opentelemetry-instrumentation-asgi==0.46b0 opentelemetry-instrumentation-fastapi==0.46b0 opentelemetry-proto==1.25.0 opentelemetry-sdk==1.25.0 opentelemetry-semantic-conventions==0.46b0 opentelemetry-util-http==0.46b0 orjson==3.10.4 overrides==7.7.0 packaging==23.2 posthog==3.5.0 protobuf==4.25.3 pyasn1==0.6.0 pyasn1_modules==0.4.0 pydantic==2.7.3 pydantic_core==2.18.4 Pygments==2.18.0 PyPika==0.48.9 pyproject_hooks==1.1.0 python-dateutil==2.9.0.post0 python-dotenv==1.0.1 python-multipart==0.0.9 PyYAML==6.0.1 regex==2024.5.15 requests==2.32.3 requests-oauthlib==2.0.0 rich==13.7.1 rsa==4.9 shellingham==1.5.4 six==1.16.0 sniffio==1.3.1 soupsieve==2.5 SQLAlchemy==2.0.30 starlette==0.37.2 sympy==1.12.1 tenacity==8.3.0 tiktoken==0.7.0 tokenizers==0.19.1 tomli==2.0.1 tqdm==4.66.4 typer==0.12.3 typing-inspect==0.9.0 typing_extensions==4.12.2 ujson==5.10.0 urllib3==2.2.1 uvicorn==0.30.1 uvloop==0.19.0 watchfiles==0.22.0 websocket-client==1.8.0 websockets==12.0 wrapt==1.16.0 yarl==1.9.4 zipp==3.19.2 |
※ pip install langchain langchain-chroma
■ create_retriever_tool 함수를 사용해 웹 문서 검색 도구를 만드는 방법을 보여준다. ▶ main.py
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 |
import os import bs4 from langgraph.checkpoint.sqlite import SqliteSaver from langchain_openai import ChatOpenAI from langchain_community.document_loaders import WebBaseLoader from langchain_text_splitters import RecursiveCharacterTextSplitter from langchain_chroma import Chroma from langchain_openai import OpenAIEmbeddings from langchain.tools.retriever import create_retriever_tool from langgraph.prebuilt import create_react_agent from langchain_core.messages import HumanMessage os.environ["OPENAI_API_KEY"] = "<OPENAI_API_KEY>" # SQLITE를 사용해 상태를 저장한다. sqliteSaver = SqliteSaver.from_conn_string(":memory:") # OpenAI 채팅 모델을 설정한다. chatOpenAI = ChatOpenAI(model = "gpt-3.5-turbo", temperature = 0) # 웹 문서를 로드한다. webBaseLoader = WebBaseLoader( web_paths = ("https://lilianweng.github.io/posts/2023-06-23-agent/",), bs_kwargs = dict(parse_only = bs4.SoupStrainer(class_ = ("post-content", "post-title", "post-header"))) ) documentList = webBaseLoader.load() # 문서를 분할한다. recursiveCharacterTextSplitter = RecursiveCharacterTextSplitter(chunk_size = 1000, chunk_overlap = 200) splitDocumentList = recursiveCharacterTextSplitter.split_documents(documentList) # 벡터 저장소를 설정한다. chroma = Chroma.from_documents(documents = splitDocumentList, embedding = OpenAIEmbeddings()) vectorStoreRetriever = chroma.as_retriever() # 검색 도구를 생성한다. tool = create_retriever_tool( vectorStoreRetriever, "BlogPostRetriever", "Searches and returns excerpts from the Autonomous Agents blog post.", ) toolList = [tool] # 대화형 에이전트를 생성한다. compiledStateGraph = create_react_agent(chatOpenAI, toolList, checkpointer = sqliteSaver) # 응답 질의를 실행한다. configurationDictionary = {"configurable" : {"thread_id" : "thread1"}} query1 = "Hi! I'm bob" for text in compiledStateGraph.stream({"messages" : [HumanMessage(content = query1)]}, config = configurationDictionary): print(text) print("----") query2 = "What is Task Decomposition?" for text in compiledStateGraph.stream({"messages" : [HumanMessage(content = query2)]}, config = configurationDictionary): print(text) print("----") query3 = "What is Task Decomposition?" for text in compiledStateGraph.stream({"messages" : [HumanMessage(content = query3)]}, config = configurationDictionary): print(text) print("----") |
▶ requirements.txt
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 |
aiohttp==3.9.5 aiosignal==1.3.1 annotated-types==0.7.0 anyio==4.4.0 asgiref==3.8.1 async-timeout==4.0.3 attrs==23.2.0 backoff==2.2.1 bcrypt==4.1.3 beautifulsoup4==4.12.3 bs4==0.0.2 build==1.2.1 cachetools==5.3.3 certifi==2024.6.2 charset-normalizer==3.3.2 chroma-hnswlib==0.7.3 chromadb==0.5.0 click==8.1.7 coloredlogs==15.0.1 dataclasses-json==0.6.7 Deprecated==1.2.14 distro==1.9.0 dnspython==2.6.1 email_validator==2.1.1 exceptiongroup==1.2.1 fastapi==0.111.0 fastapi-cli==0.0.4 filelock==3.14.0 flatbuffers==24.3.25 frozenlist==1.4.1 fsspec==2024.6.0 google-auth==2.30.0 googleapis-common-protos==1.63.1 greenlet==3.0.3 grpcio==1.64.1 h11==0.14.0 httpcore==1.0.5 httptools==0.6.1 httpx==0.27.0 huggingface-hub==0.23.3 humanfriendly==10.0 idna==3.7 importlib_metadata==7.1.0 importlib_resources==6.4.0 Jinja2==3.1.4 jsonpatch==1.33 jsonpointer==3.0.0 kubernetes==30.1.0 langchain==0.2.3 langchain-chroma==0.1.1 langchain-community==0.2.4 langchain-core==0.2.5 langchain-openai==0.1.8 langchain-text-splitters==0.2.1 langgraph==0.0.66 langsmith==0.1.77 markdown-it-py==3.0.0 MarkupSafe==2.1.5 marshmallow==3.21.3 mdurl==0.1.2 mmh3==4.1.0 monotonic==1.6 mpmath==1.3.0 multidict==6.0.5 mypy-extensions==1.0.0 numpy==1.26.4 oauthlib==3.2.2 onnxruntime==1.18.0 openai==1.33.0 opentelemetry-api==1.25.0 opentelemetry-exporter-otlp-proto-common==1.25.0 opentelemetry-exporter-otlp-proto-grpc==1.25.0 opentelemetry-instrumentation==0.46b0 opentelemetry-instrumentation-asgi==0.46b0 opentelemetry-instrumentation-fastapi==0.46b0 opentelemetry-proto==1.25.0 opentelemetry-sdk==1.25.0 opentelemetry-semantic-conventions==0.46b0 opentelemetry-util-http==0.46b0 orjson==3.10.4 overrides==7.7.0 packaging==23.2 posthog==3.5.0 protobuf==4.25.3 pyasn1==0.6.0 pyasn1_modules==0.4.0 pydantic==2.7.3 pydantic_core==2.18.4 Pygments==2.18.0 PyPika==0.48.9 pyproject_hooks==1.1.0 python-dateutil==2.9.0.post0 python-dotenv==1.0.1 python-multipart==0.0.9 PyYAML==6.0.1 regex==2024.5.15 requests==2.32.3 requests-oauthlib==2.0.0 rich==13.7.1 rsa==4.9 shellingham==1.5.4 six==1.16.0 sniffio==1.3.1 soupsieve==2.5 SQLAlchemy==2.0.30 starlette==0.37.2 sympy==1.12.1 tenacity==8.3.0 tiktoken==0.7.0 tokenizers==0.19.1 tomli==2.0.1 tqdm==4.66.4 typer==0.12.3 typing-inspect==0.9.0 typing_extensions==4.12.2 ujson==5.10.0 urllib3==2.2.1 uvicorn==0.30.1 uvloop==0.19.0 watchfiles==0.22.0 websocket-client==1.8.0 websockets==12.0 wrapt==1.16.0 yarl==1.9.4 zipp==3.19.2 |
※ pip install langchain langchain-chroma
■ 단순 대화형 RAG 애플리케이션을 만드는 방법을 보여준다. (채팅 기록 자동 업데이트) ▶ main.py
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 |
import os import bs4 from langchain_openai import ChatOpenAI from langchain_core.prompts import ChatPromptTemplate from langchain.chains.combine_documents import create_stuff_documents_chain from langchain_community.document_loaders import WebBaseLoader from langchain_text_splitters import RecursiveCharacterTextSplitter from langchain_chroma import Chroma from langchain_openai import OpenAIEmbeddings from langchain.chains import create_retrieval_chain from langchain_core.prompts import MessagesPlaceholder from langchain.chains import create_history_aware_retriever from langchain_core.chat_history import BaseChatMessageHistory from langchain_community.chat_message_histories import ChatMessageHistory from langchain_core.runnables.history import RunnableWithMessageHistory # OPANAI API 키를 설정한다. os.environ["OPENAI_API_KEY"] = "<OPENAI_API_KEY>" # OPENAI 챗봇 모델을 설정한다. chatOpenAI = ChatOpenAI(model = "gpt-3.5-turbo-0125") # 질문-응답 채팅 프롬프트 템플리트를 설정한다. questionAnswerChatPromptTemplate = ChatPromptTemplate.from_messages( [ ("system", "You are an assistant for question-answering tasks. Use the following pieces of retrieved context to answer the question. If you don't know the answer, say that you don't know. Use three sentences maximum and keep the answer concise.\n\n{context}"), MessagesPlaceholder("chatHistoryList"), ("human", "{input}"), ] ) # 질문-응답 체인을 생성한다. qnaRunnableBinding = create_stuff_documents_chain(chatOpenAI, questionAnswerChatPromptTemplate) # 웹 문서를 로딩한다. webBaseLoader = WebBaseLoader( web_paths = ("https://lilianweng.github.io/posts/2023-06-23-agent/",), bs_kwargs = dict(parse_only = bs4.SoupStrainer(class_ = ("post-content", "post-title", "post-header"))) ) documentList = webBaseLoader.load() # 문서를 분할한다. recursiveCharacterTextSplitter = RecursiveCharacterTextSplitter(chunk_size = 1000, chunk_overlap = 200) splitDocumentList = recursiveCharacterTextSplitter.split_documents(documentList) # 문서 임베딩 및 검색기를 설정한다. chroma = Chroma.from_documents(documents = splitDocumentList, embedding = OpenAIEmbeddings()) vectorStoreRetriever = chroma.as_retriever() # 질문 재구성 채팅 프롬프트 템플리트를 설정한다. contextualizeQuestionChatPromptTemplate = ChatPromptTemplate.from_messages( [ ("system", "Given a chat history and the latest user question which might reference context in the chat history, formulate a standalone question which can be understood without the chat history. Do NOT answer the question, just reformulate it if needed and otherwise return it as is."), MessagesPlaceholder("chatHistoryList"), ("human", "{input}"), ] ) # 히스토리 기반 검색기 체인을 생성한다. historyAwareRetrieverRunnableBinding = create_history_aware_retriever(chatOpenAI, vectorStoreRetriever, contextualizeQuestionChatPromptTemplate) # RAG 체인을 생성한다. ragRunnableBinding = create_retrieval_chain(historyAwareRetrieverRunnableBinding, qnaRunnableBinding) # 채팅 메시지 히스토리 딕셔너리를 설정한다. chatMessageHistoryDictionary = {} def getSessionHistory(sessionID : str) -> BaseChatMessageHistory: if sessionID not in chatMessageHistoryDictionary: chatMessageHistoryDictionary[sessionID] = ChatMessageHistory() return chatMessageHistoryDictionary[sessionID] # 대화형 RAG 실행 가능 객체를 설정한다. runnableWithMessageHistory = RunnableWithMessageHistory( ragRunnableBinding, getSessionHistory, input_messages_key = "input", history_messages_key = "chatHistoryList", output_messages_key = "answer", ) # 질문에 대한 응답 수행을 수행한다. question1 = "What is Task Decomposition?" print(f"질문 1 : {question1}") print() aiMessage1 = runnableWithMessageHistory.invoke( {"input" : question1}, config = {"configurable" : {"session_id" : "session1"}} ) print(aiMessage1["answer"]) print() question2 = "What are common ways of doing it?" print(f"질문 2 : {question2}") print() aiMessage2 = runnableWithMessageHistory.invoke( {"input" : question2}, config = {"configurable" : {"session_id" : "session1"}}, ) print(aiMessage2["answer"]) print() |
▶ requirements.txt
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 |
aiohttp==3.9.5 aiosignal==1.3.1 annotated-types==0.7.0 anyio==4.4.0 asgiref==3.8.1 async-timeout==4.0.3 attrs==23.2.0 backoff==2.2.1 bcrypt==4.1.3 beautifulsoup4==4.12.3 bs4==0.0.2 build==1.2.1 cachetools==5.3.3 certifi==2024.6.2 charset-normalizer==3.3.2 chroma-hnswlib==0.7.3 chromadb==0.5.0 click==8.1.7 coloredlogs==15.0.1 dataclasses-json==0.6.7 Deprecated==1.2.14 distro==1.9.0 dnspython==2.6.1 email_validator==2.1.1 exceptiongroup==1.2.1 fastapi==0.111.0 fastapi-cli==0.0.4 filelock==3.14.0 flatbuffers==24.3.25 frozenlist==1.4.1 fsspec==2024.6.0 google-auth==2.30.0 googleapis-common-protos==1.63.1 greenlet==3.0.3 grpcio==1.64.1 h11==0.14.0 httpcore==1.0.5 httptools==0.6.1 httpx==0.27.0 huggingface-hub==0.23.3 humanfriendly==10.0 idna==3.7 importlib_metadata==7.1.0 importlib_resources==6.4.0 Jinja2==3.1.4 jsonpatch==1.33 jsonpointer==3.0.0 kubernetes==30.1.0 langchain==0.2.3 langchain-chroma==0.1.1 langchain-community==0.2.4 langchain-core==0.2.5 langchain-openai==0.1.8 langchain-text-splitters==0.2.1 langsmith==0.1.77 markdown-it-py==3.0.0 MarkupSafe==2.1.5 marshmallow==3.21.3 mdurl==0.1.2 mmh3==4.1.0 monotonic==1.6 mpmath==1.3.0 multidict==6.0.5 mypy-extensions==1.0.0 numpy==1.26.4 oauthlib==3.2.2 onnxruntime==1.18.0 openai==1.33.0 opentelemetry-api==1.25.0 opentelemetry-exporter-otlp-proto-common==1.25.0 opentelemetry-exporter-otlp-proto-grpc==1.25.0 opentelemetry-instrumentation==0.46b0 opentelemetry-instrumentation-asgi==0.46b0 opentelemetry-instrumentation-fastapi==0.46b0 opentelemetry-proto==1.25.0 opentelemetry-sdk==1.25.0 opentelemetry-semantic-conventions==0.46b0 opentelemetry-util-http==0.46b0 orjson==3.10.4 overrides==7.7.0 packaging==23.2 posthog==3.5.0 protobuf==4.25.3 pyasn1==0.6.0 pyasn1_modules==0.4.0 pydantic==2.7.3 pydantic_core==2.18.4 Pygments==2.18.0 PyPika==0.48.9 pyproject_hooks==1.1.0 python-dateutil==2.9.0.post0 python-dotenv==1.0.1 python-multipart==0.0.9 PyYAML==6.0.1 regex==2024.5.15 requests==2.32.3 requests-oauthlib==2.0.0 rich==13.7.1 rsa==4.9 shellingham==1.5.4 six==1.16.0 sniffio==1.3.1 soupsieve==2.5 SQLAlchemy==2.0.30 starlette==0.37.2 sympy==1.12.1 tenacity==8.3.0 tiktoken==0.7.0 tokenizers==0.19.1 tomli==2.0.1 tqdm==4.66.4 typer==0.12.3 typing-inspect==0.9.0 typing_extensions==4.12.2 ujson==5.10.0 urllib3==2.2.1 uvicorn==0.30.1 uvloop==0.19.0 watchfiles==0.22.0 websocket-client==1.8.0 websockets==12.0 wrapt==1.16.0 yarl==1.9.4 zipp==3.19.2 |
※ pip install langchain
■ 단순 대화형 RAG 애플리케이션을 만드는 방법을 보여준다. (채팅 기록 수동 업데이트) ▶ main.py
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 |
import os import bs4 from langchain_openai import ChatOpenAI from langchain_core.prompts import ChatPromptTemplate from langchain_community.document_loaders import WebBaseLoader from langchain_text_splitters import RecursiveCharacterTextSplitter from langchain_chroma import Chroma from langchain_openai import OpenAIEmbeddings from langchain.chains.combine_documents import create_stuff_documents_chain from langchain.chains import create_retrieval_chain from langchain_core.prompts import MessagesPlaceholder from langchain.chains import create_history_aware_retriever from langchain_core.messages import AIMessage, HumanMessage # OPANAI API 키를 설정한다. os.environ["OPENAI_API_KEY"] = "<OPENAI_API_KEY>" # OPENAI 챗봇 모델을 설정한다. chatOpenAI = ChatOpenAI(model = "gpt-3.5-turbo-0125") # 질문-응답 채팅 프롬프트 템플리트를 설정한다. qnaChatPromptTemplate = ChatPromptTemplate.from_messages( [ ("system", "You are an assistant for question-answering tasks. Use the following pieces of retrieved context to answer the question. If you don't know the answer, say that you don't know. Use three sentences maximum and keep the answer concise.\n\n{context}"), MessagesPlaceholder("chatHistoryList"), ("human", "{input}"), ] ) # 질문-응답 체인을 생성한다. qnaRunnableBinding = create_stuff_documents_chain(chatOpenAI, qnaChatPromptTemplate) # 웹 문서를 로딩한다. webBaseLoader = WebBaseLoader( web_paths = ("https://lilianweng.github.io/posts/2023-06-23-agent/",), bs_kwargs = dict(parse_only = bs4.SoupStrainer(class_ = ("post-content", "post-title", "post-header"))) ) documentList = webBaseLoader.load() # 문서를 분할한다. recursiveCharacterTextSplitter = RecursiveCharacterTextSplitter(chunk_size = 1000, chunk_overlap = 200) splitDocumentList = recursiveCharacterTextSplitter.split_documents(documentList) # 문서 임베딩 및 검색기를 설정한다. chroma = Chroma.from_documents(documents = splitDocumentList, embedding = OpenAIEmbeddings()) vectorStoreRetriever = chroma.as_retriever() # 질문 재구성 채팅 프롬프트 템플리트를 설정한다. contextualizeQuestionChatPromptTemplate = ChatPromptTemplate.from_messages( [ ("system", "Given a chat history and the latest user question which might reference context in the chat history, formulate a standalone question which can be understood without the chat history. Do NOT answer the question, just reformulate it if needed and otherwise return it as is."), MessagesPlaceholder("chatHistoryList"), ("human", "{input}"), ] ) # 히스토리 기반 검색기 체인을 생성한다. historyAwareRetrieverRunnableBinding = create_history_aware_retriever(chatOpenAI, vectorStoreRetriever, contextualizeQuestionChatPromptTemplate) # RAG 체인을 생성한다. ragRunnableBinding = create_retrieval_chain(historyAwareRetrieverRunnableBinding, qnaRunnableBinding) # 질문-응답을 실행한다. chatHistoryList = [] question1 = "What is Task Decomposition?" print(f"질문 1 : {question1}") print() aiMessage1 = ragRunnableBinding.invoke({"input" : question1, "chatHistoryList" : chatHistoryList}) print(aiMessage1["answer"]) print() chatHistoryList.extend( [ HumanMessage(content = question1), AIMessage(content = aiMessage1["answer"]), ] ) question2 = "What are common ways of doing it?" print(f"질문 2 : {question1}") print() aiMessage2 = ragRunnableBinding.invoke({"input" : question2, "chatHistoryList" : chatHistoryList}) print(aiMessage2["answer"]) print() """ 질문 1 : What is Task Decomposition? Task decomposition involves breaking down complex tasks into smaller and simpler steps to assist an agent in planning and executing the task efficiently. Techniques like Chain of Thought (CoT) and Tree of Thoughts can be employed to decompose tasks into manageable subtasks, enhancing the model's performance on complex tasks. Task decomposition can be facilitated through simple prompting, task-specific instructions, or human inputs to guide the agent through the sequential execution of the task. 질문 2 : What is Task Decomposition? Task decomposition can be achieved through various methods, including using language models with simple prompts, task-specific instructions, or human inputs to break down complex tasks into manageable subtasks. Additionally, techniques like the Tree of Thoughts approach can explore multiple reasoning possibilities at each step, creating a tree structure of thoughts to guide the agent through the task. Different search processes, such as BFS or DFS, can be utilized to evaluate states during the task decomposition process. """ |
▶ requirements.txt
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 |
aiohttp==3.9.5 aiosignal==1.3.1 annotated-types==0.7.0 anyio==4.4.0 asgiref==3.8.1 async-timeout==4.0.3 attrs==23.2.0 backoff==2.2.1 bcrypt==4.1.3 beautifulsoup4==4.12.3 bs4==0.0.2 build==1.2.1 cachetools==5.3.3 certifi==2024.6.2 charset-normalizer==3.3.2 chroma-hnswlib==0.7.3 chromadb==0.5.0 click==8.1.7 coloredlogs==15.0.1 dataclasses-json==0.6.7 Deprecated==1.2.14 distro==1.9.0 dnspython==2.6.1 email_validator==2.1.1 exceptiongroup==1.2.1 fastapi==0.111.0 fastapi-cli==0.0.4 filelock==3.14.0 flatbuffers==24.3.25 frozenlist==1.4.1 fsspec==2024.6.0 google-auth==2.30.0 googleapis-common-protos==1.63.1 greenlet==3.0.3 grpcio==1.64.1 h11==0.14.0 httpcore==1.0.5 httptools==0.6.1 httpx==0.27.0 huggingface-hub==0.23.3 humanfriendly==10.0 idna==3.7 importlib_metadata==7.1.0 importlib_resources==6.4.0 Jinja2==3.1.4 jsonpatch==1.33 jsonpointer==3.0.0 kubernetes==30.1.0 langchain==0.2.3 langchain-chroma==0.1.1 langchain-community==0.2.4 langchain-core==0.2.5 langchain-openai==0.1.8 langchain-text-splitters==0.2.1 langsmith==0.1.77 markdown-it-py==3.0.0 MarkupSafe==2.1.5 marshmallow==3.21.3 mdurl==0.1.2 mmh3==4.1.0 monotonic==1.6 mpmath==1.3.0 multidict==6.0.5 mypy-extensions==1.0.0 numpy==1.26.4 oauthlib==3.2.2 onnxruntime==1.18.0 openai==1.33.0 opentelemetry-api==1.25.0 opentelemetry-exporter-otlp-proto-common==1.25.0 opentelemetry-exporter-otlp-proto-grpc==1.25.0 opentelemetry-instrumentation==0.46b0 opentelemetry-instrumentation-asgi==0.46b0 opentelemetry-instrumentation-fastapi==0.46b0 opentelemetry-proto==1.25.0 opentelemetry-sdk==1.25.0 opentelemetry-semantic-conventions==0.46b0 opentelemetry-util-http==0.46b0 orjson==3.10.4 overrides==7.7.0 packaging==23.2 posthog==3.5.0 protobuf==4.25.3 pyasn1==0.6.0 pyasn1_modules==0.4.0 pydantic==2.7.3 pydantic_core==2.18.4 Pygments==2.18.0 PyPika==0.48.9 pyproject_hooks==1.1.0 python-dateutil==2.9.0.post0 python-dotenv==1.0.1 python-multipart==0.0.9 PyYAML==6.0.1 regex==2024.5.15 requests==2.32.3 requests-oauthlib==2.0.0 rich==13.7.1 rsa==4.9 shellingham==1.5.4 six==1.16.0 sniffio==1.3.1 soupsieve==2.5 SQLAlchemy==2.0.30 starlette==0.37.2 sympy==1.12.1 tenacity==8.3.0 tiktoken==0.7.0 tokenizers==0.19.1 tomli==2.0.1 tqdm==4.66.4 typer==0.12.3 typing-inspect==0.9.0 typing_extensions==4.12.2 ujson==5.10.0 urllib3==2.2.1 uvicorn==0.30.1 uvloop==0.19.0 watchfiles==0.22.0 websocket-client==1.8.0 websockets==12.0 wrapt==1.16.0 yarl==1.9.4 zipp==3.19.2 |
※ pip install langchain
■ create_history_aware_retriever 함수를 사용해 히스토리 기반 검색기를 생성하는 방법을 보여준다. ▶ main.py
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 |
import os import bs4 from langchain_community.document_loaders import WebBaseLoader from langchain_text_splitters import RecursiveCharacterTextSplitter from langchain_chroma import Chroma from langchain_openai import OpenAIEmbeddings from langchain_openai import ChatOpenAI from langchain_core.prompts import ChatPromptTemplate from langchain.chains.combine_documents import create_stuff_documents_chain from langchain_core.prompts import MessagesPlaceholder from langchain.chains import create_history_aware_retriever # OPANAI API 키를 설정한다. os.environ["OPENAI_API_KEY"] = "<OPENAI_API_KEY>" # OPENAI 챗봇 모델을 설정한다. chatOpenAI = ChatOpenAI(model = "gpt-3.5-turbo-0125") # 질문-응답 채팅 프롬프트 템플리트를 설정한다. questionAnswerChatPromptTemplate = ChatPromptTemplate.from_messages( [ ("system", "You are an assistant for question-answering tasks. Use the following pieces of retrieved context to answer the question. If you don't know the answer, say that you don't know. Use three sentences maximum and keep the answer concise.\n\n{context}"), MessagesPlaceholder("chatHistoryList"), ("human", "{input}"), ] ) # 질문-응답 체인을 생성한다. questionAnswerRunnableBinding = create_stuff_documents_chain(chatOpenAI, questionAnswerChatPromptTemplate) # 웹 문서를 로딩한다. webBaseLoader = WebBaseLoader( web_paths = ("https://lilianweng.github.io/posts/2023-06-23-agent/",), bs_kwargs = dict(parse_only = bs4.SoupStrainer(class_ = ("post-content", "post-title", "post-header"))) ) documentList = webBaseLoader.load() # 문서를 분할한다. recursiveCharacterTextSplitter = RecursiveCharacterTextSplitter(chunk_size = 1000, chunk_overlap = 200) splitDocumentList = recursiveCharacterTextSplitter.split_documents(documentList) # 문서 임베딩 및 검색기를 설정한다. chroma = Chroma.from_documents(documents = splitDocumentList, embedding = OpenAIEmbeddings()) vectorStoreRetriever = chroma.as_retriever() # 질문 재구성 채팅 프롬프트 템플리트를 설정한다. contextualizeQuestionChatPromptTemplate = ChatPromptTemplate.from_messages( [ ("system", "Given a chat history and the latest user question which might reference context in the chat history, formulate a standalone question which can be understood without the chat history. Do NOT answer the question, just reformulate it if needed and otherwise return it as is."), MessagesPlaceholder("chatHistoryList"), ("human", "{input}"), ] ) # 히스토리 기반 검색기 체인을 생성한다. historyAwareRetrieverRunnableBinding = create_history_aware_retriever(chatOpenAI, vectorStoreRetriever, contextualizeQuestionChatPromptTemplate) |
▶ requirements.txt
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 |
aiohttp==3.9.5 aiosignal==1.3.1 annotated-types==0.7.0 anyio==4.4.0 asgiref==3.8.1 async-timeout==4.0.3 attrs==23.2.0 backoff==2.2.1 bcrypt==4.1.3 beautifulsoup4==4.12.3 bs4==0.0.2 build==1.2.1 cachetools==5.3.3 certifi==2024.6.2 charset-normalizer==3.3.2 chroma-hnswlib==0.7.3 chromadb==0.5.0 click==8.1.7 coloredlogs==15.0.1 dataclasses-json==0.6.7 Deprecated==1.2.14 distro==1.9.0 dnspython==2.6.1 email_validator==2.1.1 exceptiongroup==1.2.1 fastapi==0.111.0 fastapi-cli==0.0.4 filelock==3.14.0 flatbuffers==24.3.25 frozenlist==1.4.1 fsspec==2024.6.0 google-auth==2.30.0 googleapis-common-protos==1.63.1 greenlet==3.0.3 grpcio==1.64.1 h11==0.14.0 httpcore==1.0.5 httptools==0.6.1 httpx==0.27.0 huggingface-hub==0.23.3 humanfriendly==10.0 idna==3.7 importlib_metadata==7.1.0 importlib_resources==6.4.0 Jinja2==3.1.4 jsonpatch==1.33 jsonpointer==3.0.0 kubernetes==30.1.0 langchain==0.2.3 langchain-chroma==0.1.1 langchain-community==0.2.4 langchain-core==0.2.5 langchain-openai==0.1.8 langchain-text-splitters==0.2.1 langsmith==0.1.77 markdown-it-py==3.0.0 MarkupSafe==2.1.5 marshmallow==3.21.3 mdurl==0.1.2 mmh3==4.1.0 monotonic==1.6 mpmath==1.3.0 multidict==6.0.5 mypy-extensions==1.0.0 numpy==1.26.4 oauthlib==3.2.2 onnxruntime==1.18.0 openai==1.33.0 opentelemetry-api==1.25.0 opentelemetry-exporter-otlp-proto-common==1.25.0 opentelemetry-exporter-otlp-proto-grpc==1.25.0 opentelemetry-instrumentation==0.46b0 opentelemetry-instrumentation-asgi==0.46b0 opentelemetry-instrumentation-fastapi==0.46b0 opentelemetry-proto==1.25.0 opentelemetry-sdk==1.25.0 opentelemetry-semantic-conventions==0.46b0 opentelemetry-util-http==0.46b0 orjson==3.10.4 overrides==7.7.0 packaging==23.2 posthog==3.5.0 protobuf==4.25.3 pyasn1==0.6.0 pyasn1_modules==0.4.0 pydantic==2.7.3 pydantic_core==2.18.4 Pygments==2.18.0 PyPika==0.48.9 pyproject_hooks==1.1.0 python-dateutil==2.9.0.post0 python-dotenv==1.0.1 python-multipart==0.0.9 PyYAML==6.0.1 regex==2024.5.15 requests==2.32.3 requests-oauthlib==2.0.0 rich==13.7.1 rsa==4.9 shellingham==1.5.4 six==1.16.0 sniffio==1.3.1 soupsieve==2.5 SQLAlchemy==2.0.30 starlette==0.37.2 sympy==1.12.1 tenacity==8.3.0 tiktoken==0.7.0 tokenizers==0.19.1 tomli==2.0.1 tqdm==4.66.4 typer==0.12.3 typing-inspect==0.9.0 typing_extensions==4.12.2 ujson==5.10.0 urllib3==2.2.1 uvicorn==0.30.1 uvloop==0.19.0 watchfiles==0.22.0 websocket-client==1.8.0 websockets==12.0 wrapt==1.16.0 yarl==1.9.4 zipp==3.19.2 |
※ pip install langchain langchain-chroma langchain-community
■ create_retrieval_chain 함수를 사용해 문서 검색 모듈과 질문/답변 생성 모듈을 연결하는 방법을 보여준다. ▶ main.py
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 |
import os import bs4 from langchain_community.document_loaders import WebBaseLoader from langchain_text_splitters import RecursiveCharacterTextSplitter from langchain_chroma import Chroma from langchain_openai import OpenAIEmbeddings from langchain_openai import ChatOpenAI from langchain_core.prompts import ChatPromptTemplate from langchain.chains.combine_documents import create_stuff_documents_chain from langchain.chains import create_retrieval_chain os.environ["OPENAI_API_KEY"] = "<OPENAI_API_KEY>" webBaseLoader = WebBaseLoader( web_paths = ("https://lilianweng.github.io/posts/2023-06-23-agent/",), bs_kwargs = dict(parse_only = bs4.SoupStrainer(class_ = ("media_end_head_headline", "newsct_body"))) ) documentList = webBaseLoader.load() recursiveCharacterTextSplitter = RecursiveCharacterTextSplitter(chunk_size = 1000, chunk_overlap = 200) splitDocumentList = recursiveCharacterTextSplitter.split_documents(documentList) chroma = Chroma.from_documents(documents = splitDocumentList, embedding = OpenAIEmbeddings()) vectorStoreRetriever = chroma.as_retriever() systemString = ( "You are an assistant for question-answering tasks. " "Use the following pieces of retrieved context to answer " "the question. If you don't know the answer, say that you " "don't know. Use three sentences maximum and keep the " "answer concise." "\n\n" "{context}" ) chatPromptTemplate = ChatPromptTemplate.from_messages( [ ("system", systemString), ("human", "{input}"), ] ) chatOpenAI = ChatOpenAI(model = "gpt-3.5-turbo-0125") questionAnswerRunnableBinding = create_stuff_documents_chain(chatOpenAI, chatPromptTemplate) ragRunnableBinding = create_retrieval_chain(vectorStoreRetriever, questionAnswerRunnableBinding) responseDictionary = ragRunnableBinding.invoke({"input" : "What is Task Decomposition?"}) print(responseDictionary["answer"]) """ Task decomposition involves breaking down a complex task into smaller and more manageable subtasks. This process helps agents or models tackle difficult tasks step by step, making it easier to plan and execute the overall task efficiently. Task decomposition can be achieved through techniques like Chain of Thought or Tree of Thoughts, which guide the model in breaking down tasks into simpler components. """ |
▶ requirements.txt
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 |
aiohttp==3.9.5 aiosignal==1.3.1 annotated-types==0.7.0 anyio==4.4.0 asgiref==3.8.1 async-timeout==4.0.3 attrs==23.2.0 backoff==2.2.1 bcrypt==4.1.3 beautifulsoup4==4.12.3 bs4==0.0.2 build==1.2.1 cachetools==5.3.3 certifi==2024.6.2 charset-normalizer==3.3.2 chroma-hnswlib==0.7.3 chromadb==0.5.0 click==8.1.7 coloredlogs==15.0.1 dataclasses-json==0.6.7 Deprecated==1.2.14 distro==1.9.0 dnspython==2.6.1 email_validator==2.1.1 exceptiongroup==1.2.1 fastapi==0.111.0 fastapi-cli==0.0.4 filelock==3.14.0 flatbuffers==24.3.25 frozenlist==1.4.1 fsspec==2024.6.0 google-auth==2.30.0 googleapis-common-protos==1.63.1 greenlet==3.0.3 grpcio==1.64.1 h11==0.14.0 httpcore==1.0.5 httptools==0.6.1 httpx==0.27.0 huggingface-hub==0.23.3 humanfriendly==10.0 idna==3.7 importlib_metadata==7.1.0 importlib_resources==6.4.0 Jinja2==3.1.4 jsonpatch==1.33 jsonpointer==3.0.0 kubernetes==30.1.0 langchain==0.2.3 langchain-chroma==0.1.1 langchain-community==0.2.4 langchain-core==0.2.5 langchain-openai==0.1.8 langchain-text-splitters==0.2.1 langsmith==0.1.77 markdown-it-py==3.0.0 MarkupSafe==2.1.5 marshmallow==3.21.3 mdurl==0.1.2 mmh3==4.1.0 monotonic==1.6 mpmath==1.3.0 multidict==6.0.5 mypy-extensions==1.0.0 numpy==1.26.4 oauthlib==3.2.2 onnxruntime==1.18.0 openai==1.33.0 opentelemetry-api==1.25.0 opentelemetry-exporter-otlp-proto-common==1.25.0 opentelemetry-exporter-otlp-proto-grpc==1.25.0 opentelemetry-instrumentation==0.46b0 opentelemetry-instrumentation-asgi==0.46b0 opentelemetry-instrumentation-fastapi==0.46b0 opentelemetry-proto==1.25.0 opentelemetry-sdk==1.25.0 opentelemetry-semantic-conventions==0.46b0 opentelemetry-util-http==0.46b0 orjson==3.10.4 overrides==7.7.0 packaging==23.2 posthog==3.5.0 protobuf==4.25.3 pyasn1==0.6.0 pyasn1_modules==0.4.0 pydantic==2.7.3 pydantic_core==2.18.4 Pygments==2.18.0 PyPika==0.48.9 pyproject_hooks==1.1.0 python-dateutil==2.9.0.post0 python-dotenv==1.0.1 python-multipart==0.0.9 PyYAML==6.0.1 regex==2024.5.15 requests==2.32.3 requests-oauthlib==2.0.0 rich==13.7.1 rsa==4.9 shellingham==1.5.4 six==1.16.0 sniffio==1.3.1 soupsieve==2.5 SQLAlchemy==2.0.30 starlette==0.37.2 sympy==1.12.1 tenacity==8.3.0 tiktoken==0.7.0 tokenizers==0.19.1 tomli==2.0.1 tqdm==4.66.4 typer==0.12.3 typing-inspect==0.9.0 typing_extensions==4.12.2 ujson==5.10.0 urllib3==2.2.1 uvicorn==0.30.1 uvloop==0.19.0 watchfiles==0.22.0 websocket-client==1.8.0 websockets==12.0 wrapt==1.16.0 yarl==1.9.4 zipp==3.19.2 |
※ pip install
■ Chroma 클래스의 as_retriever 메소드를 사용해 VectorStoreRetriever 객체를 만드는 방법을 보여준다. ▶ main.py
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 |
import os import bs4 from langchain_community.document_loaders import WebBaseLoader from langchain_text_splitters import RecursiveCharacterTextSplitter from langchain_chroma import Chroma from langchain_openai import OpenAIEmbeddings os.environ["OPENAI_API_KEY"] = "<OPENAI_API_KEY>" webBaseLoader = WebBaseLoader( web_paths = ("https://lilianweng.github.io/posts/2023-06-23-agent/",), bs_kwargs = dict(parse_only = bs4.SoupStrainer(class_ = ("post-content", "post-title", "post-header"))) ) documentList = webBaseLoader.load() recursiveCharacterTextSplitter = RecursiveCharacterTextSplitter(chunk_size = 1000, chunk_overlap = 200) splitDocumentList = recursiveCharacterTextSplitter.split_documents(documentList) chroma = Chroma.from_documents(documents = splitDocumentList, embedding = OpenAIEmbeddings()) vectorStoreRetriever = chroma.as_retriever() |
▶ requirements.txt
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 |
aiohttp==3.9.5 aiosignal==1.3.1 annotated-types==0.7.0 anyio==4.4.0 asgiref==3.8.1 async-timeout==4.0.3 attrs==23.2.0 backoff==2.2.1 bcrypt==4.1.3 beautifulsoup4==4.12.3 bs4==0.0.2 build==1.2.1 cachetools==5.3.3 certifi==2024.6.2 charset-normalizer==3.3.2 chroma-hnswlib==0.7.3 chromadb==0.5.0 click==8.1.7 coloredlogs==15.0.1 dataclasses-json==0.6.7 Deprecated==1.2.14 distro==1.9.0 dnspython==2.6.1 email_validator==2.1.1 exceptiongroup==1.2.1 fastapi==0.111.0 fastapi-cli==0.0.4 filelock==3.14.0 flatbuffers==24.3.25 frozenlist==1.4.1 fsspec==2024.6.0 google-auth==2.30.0 googleapis-common-protos==1.63.1 greenlet==3.0.3 grpcio==1.64.1 h11==0.14.0 httpcore==1.0.5 httptools==0.6.1 httpx==0.27.0 huggingface-hub==0.23.3 humanfriendly==10.0 idna==3.7 importlib_metadata==7.1.0 importlib_resources==6.4.0 Jinja2==3.1.4 jsonpatch==1.33 jsonpointer==3.0.0 kubernetes==30.1.0 langchain==0.2.3 langchain-chroma==0.1.1 langchain-community==0.2.4 langchain-core==0.2.5 langchain-openai==0.1.8 langchain-text-splitters==0.2.1 langsmith==0.1.77 markdown-it-py==3.0.0 MarkupSafe==2.1.5 marshmallow==3.21.3 mdurl==0.1.2 mmh3==4.1.0 monotonic==1.6 mpmath==1.3.0 multidict==6.0.5 mypy-extensions==1.0.0 numpy==1.26.4 oauthlib==3.2.2 onnxruntime==1.18.0 openai==1.33.0 opentelemetry-api==1.25.0 opentelemetry-exporter-otlp-proto-common==1.25.0 opentelemetry-exporter-otlp-proto-grpc==1.25.0 opentelemetry-instrumentation==0.46b0 opentelemetry-instrumentation-asgi==0.46b0 opentelemetry-instrumentation-fastapi==0.46b0 opentelemetry-proto==1.25.0 opentelemetry-sdk==1.25.0 opentelemetry-semantic-conventions==0.46b0 opentelemetry-util-http==0.46b0 orjson==3.10.4 overrides==7.7.0 packaging==23.2 posthog==3.5.0 protobuf==4.25.3 pyasn1==0.6.0 pyasn1_modules==0.4.0 pydantic==2.7.3 pydantic_core==2.18.4 Pygments==2.18.0 PyPika==0.48.9 pyproject_hooks==1.1.0 python-dateutil==2.9.0.post0 python-dotenv==1.0.1 python-multipart==0.0.9 PyYAML==6.0.1 regex==2024.5.15 requests==2.32.3 requests-oauthlib==2.0.0 rich==13.7.1 rsa==4.9 shellingham==1.5.4 six==1.16.0 sniffio==1.3.1 soupsieve==2.5 SQLAlchemy==2.0.30 starlette==0.37.2 sympy==1.12.1 tenacity==8.3.0 tiktoken==0.7.0 tokenizers==0.19.1 tomli==2.0.1 tqdm==4.66.4 typer==0.12.3 typing-inspect==0.9.0 typing_extensions==4.12.2 ujson==5.10.0 urllib3==2.2.1 uvicorn==0.30.1 uvloop==0.19.0 watchfiles==0.22.0 websocket-client==1.8.0 websockets==12.0 wrapt==1.16.0 yarl==1.9.4 zipp==3.19.2 |
※ pip install langchain langchain-chroma
■ RecursiveCharacterTextSplitter 클래스의 split_documents 메소드를 사용해 문서를 분할하는 방법을 보여준다. ▶ main.py
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 |
import bs4 from langchain_community.document_loaders import WebBaseLoader from langchain_text_splitters import RecursiveCharacterTextSplitter webBaseLoader = WebBaseLoader( web_paths = ("https://lilianweng.github.io/posts/2023-06-23-agent/",), bs_kwargs = dict(parse_only = bs4.SoupStrainer(class_ = ("post-content", "post-title", "post-header"))) ) documentList = webBaseLoader.load() recursiveCharacterTextSplitter = RecursiveCharacterTextSplitter(chunk_size = 1000, chunk_overlap = 200) splitDocumentList = recursiveCharacterTextSplitter.split_documents(documentList) print(f"문서 리스트 요소 수 : {len(documentList )}") print(f"분할 문서 리스트 요소 수 : {len(splitDocumentList)}") """ 문서 리스트 요소 수 : 1 분할 문서 리스트 요소 수 : 66 """ |
▶ requirements.txt
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 |
aiohttp==3.9.5 aiosignal==1.3.1 annotated-types==0.7.0 async-timeout==4.0.3 attrs==23.2.0 beautifulsoup4==4.12.3 bs4==0.0.2 certifi==2024.6.2 charset-normalizer==3.3.2 dataclasses-json==0.6.7 frozenlist==1.4.1 greenlet==3.0.3 idna==3.7 jsonpatch==1.33 jsonpointer==3.0.0 langchain==0.2.3 langchain-community==0.2.4 langchain-core==0.2.5 langchain-text-splitters==0.2.1 langsmith==0.1.76 marshmallow==3.21.3 multidict==6.0.5 mypy-extensions==1.0.0 numpy==1.26.4 orjson==3.10.4 packaging==23.2 pydantic==2.7.3 pydantic_core==2.18.4 PyYAML==6.0.1 requests==2.32.3 soupsieve==2.5 SQLAlchemy==2.0.30 tenacity==8.3.0 typing-inspect==0.9.0 typing_extensions==4.12.2 urllib3==2.2.1 yarl==1.9.4 |
※ pip install bs4 langchain_community 명령을
■ WebBaseLoader 클래스의 load 메소드를 사용해 웹 문서를 로드하는 방법을 보여준다. ▶ main.py
|
1 2 3 4 5 6 7 8 9 10 11 12 |
import bs4 from langchain_community.document_loaders import WebBaseLoader webBaseLoader = WebBaseLoader( web_paths = ("https://lilianweng.github.io/posts/2023-06-23-agent/",), bs_kwargs = dict(parse_only = bs4.SoupStrainer(class_ = ("post-content", "post-title", "post-header"))) ) documentList = webBaseLoader.load() |
▶ requirements.txt
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 |
aiohttp==3.9.5 aiosignal==1.3.1 annotated-types==0.7.0 async-timeout==4.0.3 attrs==23.2.0 beautifulsoup4==4.12.3 bs4==0.0.2 certifi==2024.6.2 charset-normalizer==3.3.2 dataclasses-json==0.6.7 frozenlist==1.4.1 greenlet==3.0.3 idna==3.7 jsonpatch==1.33 jsonpointer==3.0.0 langchain==0.2.3 langchain-community==0.2.4 langchain-core==0.2.5 langchain-text-splitters==0.2.1 langsmith==0.1.76 marshmallow==3.21.3 multidict==6.0.5 mypy-extensions==1.0.0 numpy==1.26.4 orjson==3.10.4 packaging==23.2 pydantic==2.7.3 pydantic_core==2.18.4 PyYAML==6.0.1 requests==2.32.3 soupsieve==2.5 SQLAlchemy==2.0.30 tenacity==8.3.0 typing-inspect==0.9.0 typing_extensions==4.12.2 urllib3==2.2.1 yarl==1.9.4 |
※ pip install bs4 langchain_community
■ 단순 검색 증강 생성(RAG, Retrieval Augmented Generation) 애플리케이션을 만드는 방법을 보여준다. ▶ main.py
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 |
import bs4 import os from langchain_community.document_loaders import WebBaseLoader from langchain_text_splitters import RecursiveCharacterTextSplitter from langchain_chroma import Chroma from langchain_openai import OpenAIEmbeddings from langchain import hub from langchain_core.runnables import RunnablePassthrough from langchain_openai import ChatOpenAI from langchain_core.output_parsers import StrOutputParser os.environ["OPENAI_API_KEY"] = "<OPENAI_API_KEY>" # 웹 문서를 로드한다. webBaseLoader = WebBaseLoader( web_paths = ("https://lilianweng.github.io/posts/2023-06-23-agent/",), bs_kwargs = dict(parse_only = bs4.SoupStrainer(class_ = ("post-content", "post-title", "post-header"))) ) documentList = webBaseLoader.load() # 문서 리스트 : 1건 # 문서를 분할한다. recursiveCharacterTextSplitter = RecursiveCharacterTextSplitter(chunk_size = 1000, chunk_overlap = 200) splitDocumentList = recursiveCharacterTextSplitter.split_documents(documentList) # 분할 문서 리스트 : 66건 # 분할한 문서를 벡터 저장소에 저장한다. chroma = Chroma.from_documents(documents = splitDocumentList, embedding = OpenAIEmbeddings()) vectorStoreRetriever = chroma.as_retriever() # 벡터 저장소 검색기 # 문서 내용 병합 함수를 정의한다. def mergeDocumentContent(documentList): return "\n\n".join(document.page_content for document in documentList) # 랭체인 허브에서 프롬프트 템플리트를 가져온다. chatPromptTemplate = hub.pull("rlm/rag-prompt") # 질의응답 모델을 설정한다. chatOpenAI = ChatOpenAI(model = "gpt-3.5-turbo-0125") # 질의응답 체인을 설정한다. runnableSequence = ( {"context" : vectorStoreRetriever | mergeDocumentContent, "question" : RunnablePassthrough()} | chatPromptTemplate | chatOpenAI | StrOutputParser() ) # 질의응답을 실행한다. resultString = runnableSequence.invoke("What is Task Decomposition?") print(resultString) """ Task Decomposition is a technique that breaks down complex tasks into smaller and simpler steps to make them more manageable. It involves transforming big tasks into multiple smaller tasks to enhance model performance. This process can be done using prompting techniques, task-specific instructions, or with human inputs. """ |
▶ requirements.txt
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 |
aiohttp==3.9.5 aiosignal==1.3.1 annotated-types==0.7.0 anyio==4.4.0 asgiref==3.8.1 async-timeout==4.0.3 attrs==23.2.0 backoff==2.2.1 bcrypt==4.1.3 beautifulsoup4==4.12.3 bs4==0.0.2 build==1.2.1 cachetools==5.3.3 certifi==2024.6.2 charset-normalizer==3.3.2 chroma-hnswlib==0.7.3 chromadb==0.5.0 click==8.1.7 coloredlogs==15.0.1 dataclasses-json==0.6.7 Deprecated==1.2.14 distro==1.9.0 dnspython==2.6.1 email_validator==2.1.1 exceptiongroup==1.2.1 fastapi==0.111.0 fastapi-cli==0.0.4 filelock==3.14.0 flatbuffers==24.3.25 frozenlist==1.4.1 fsspec==2024.6.0 google-auth==2.30.0 googleapis-common-protos==1.63.1 greenlet==3.0.3 grpcio==1.64.1 h11==0.14.0 httpcore==1.0.5 httptools==0.6.1 httpx==0.27.0 huggingface-hub==0.23.3 humanfriendly==10.0 idna==3.7 importlib_metadata==7.1.0 importlib_resources==6.4.0 Jinja2==3.1.4 jsonpatch==1.33 jsonpointer==3.0.0 kubernetes==30.1.0 langchain==0.2.3 langchain-chroma==0.1.1 langchain-community==0.2.4 langchain-core==0.2.5 langchain-openai==0.1.8 langchain-text-splitters==0.2.1 langchainhub==0.1.18 langsmith==0.1.76 markdown-it-py==3.0.0 MarkupSafe==2.1.5 marshmallow==3.21.3 mdurl==0.1.2 mmh3==4.1.0 monotonic==1.6 mpmath==1.3.0 multidict==6.0.5 mypy-extensions==1.0.0 numpy==1.26.4 oauthlib==3.2.2 onnxruntime==1.18.0 openai==1.33.0 opentelemetry-api==1.25.0 opentelemetry-exporter-otlp-proto-common==1.25.0 opentelemetry-exporter-otlp-proto-grpc==1.25.0 opentelemetry-instrumentation==0.46b0 opentelemetry-instrumentation-asgi==0.46b0 opentelemetry-instrumentation-fastapi==0.46b0 opentelemetry-proto==1.25.0 opentelemetry-sdk==1.25.0 opentelemetry-semantic-conventions==0.46b0 opentelemetry-util-http==0.46b0 orjson==3.10.4 overrides==7.7.0 packaging==23.2 posthog==3.5.0 protobuf==4.25.3 pyasn1==0.6.0 pyasn1_modules==0.4.0 pydantic==2.7.3 pydantic_core==2.18.4 Pygments==2.18.0 PyPika==0.48.9 pyproject_hooks==1.1.0 python-dateutil==2.9.0.post0 python-dotenv==1.0.1 python-multipart==0.0.9 PyYAML==6.0.1 regex==2024.5.15 requests==2.32.3 requests-oauthlib==2.0.0 rich==13.7.1 rsa==4.9 shellingham==1.5.4 six==1.16.0 sniffio==1.3.1 soupsieve==2.5 SQLAlchemy==2.0.30 starlette==0.37.2 sympy==1.12.1 tenacity==8.3.0 tiktoken==0.7.0 tokenizers==0.19.1 tomli==2.0.1 tqdm==4.66.4 typer==0.12.3 types-requests==2.32.0.20240602 typing-inspect==0.9.0 typing_extensions==4.12.2 ujson==5.10.0 urllib3==2.2.1 uvicorn==0.30.1 uvloop==0.19.0 watchfiles==0.22.0 websocket-client==1.8.0 websockets==12.0 wrapt==1.16.0 yarl==1.9.4 zipp==3.19.2 |
※ pip install bs4
■ bs4 패키지를 설치하는 방법을 보여준다. 1. 명령 프롬프트를 실행한다. 2. 명령 프롬프트에서 아래 명령을 실행한다. ▶ 실행 명령
|
1 2 3 |
pip bs4 |