Retrieval-erweiterte Generierung (RAG) mit Milvus und LangChain

Diese Anleitung zeigt, wie man ein Retrieval-Augmented Generation (RAG) System mit LangChain und Milvus aufbaut.

Das RAG-System kombiniert ein Retrieval-System mit einem generativen Modell, um neuen Text auf der Grundlage einer vorgegebenen Aufforderung zu generieren. Das System ruft zunächst relevante Dokumente aus einem Korpus mit Milvus ab und verwendet dann ein generatives Modell, um neuen Text auf der Grundlage der abgerufenen Dokumente zu erzeugen.

LangChain ist ein Framework für die Entwicklung von Anwendungen, die auf großen Sprachmodellen (LLMs) basieren. Milvus ist die weltweit fortschrittlichste Open-Source-Vektordatenbank, die für die Einbettung von Ähnlichkeitssuche und KI-Anwendungen entwickelt wurde.

Voraussetzungen

Vergewissern Sie sich, dass Sie die folgenden Abhängigkeiten installiert haben, bevor Sie dieses Notizbuch ausführen:

$ pip install --upgrade --quiet  langchain langchain-core langchain-community langchain-text-splitters langchain-milvus langchain-openai bs4

Wir werden die Modelle von OpenAI verwenden. Sie sollten den Api-Schlüssel OPENAI_API_KEY als Umgebungsvariable vorbereiten.

import os

os.environ["OPENAI_API_KEY"] = "sk-***********"

Bereiten Sie die Daten vor

Wir verwenden den Langchain WebBaseLoader, um Dokumente aus Webquellen zu laden und sie mit dem RecursiveCharacterTextSplitter in Stücke zu zerlegen.

import bs4
from langchain_community.document_loaders import WebBaseLoader
from langchain_text_splitters import RecursiveCharacterTextSplitter

# Create a WebBaseLoader instance to load documents from web sources
loader = WebBaseLoader(
    web_paths=(
        "https://lilianweng.github.io/posts/2023-06-23-agent/",
        "https://lilianweng.github.io/posts/2023-03-15-prompt-engineering/",
    ),
    bs_kwargs=dict(
        parse_only=bs4.SoupStrainer(
            class_=("post-content", "post-title", "post-header")
        )
    ),
)
# Load documents from web sources using the loader
documents = loader.load()
# Initialize a RecursiveCharacterTextSplitter for splitting text into chunks
text_splitter = RecursiveCharacterTextSplitter(chunk_size=2000, chunk_overlap=200)

# Split the documents into chunks using the text_splitter
docs = text_splitter.split_documents(documents)

# Let's take a look at the first document
docs[1]

Document(page_content='Fig. 1. Overview of a LLM-powered autonomous agent system.\nComponent One: Planning#\nA complicated task usually involves many steps. An agent needs to know what they are and plan ahead.\nTask Decomposition#\nChain of thought (CoT; Wei et al. 2022) has become a standard prompting technique for enhancing model performance on complex tasks. The model is instructed to “think step by step” to utilize more test-time computation to decompose hard tasks into smaller and simpler steps. CoT transforms big tasks into multiple manageable tasks and shed lights into an interpretation of the model’s thinking process.\nTree of Thoughts (Yao et al. 2023) extends CoT by exploring multiple reasoning possibilities at each step. It first decomposes the problem into multiple thought steps and generates multiple thoughts per step, creating a tree structure. The search process can be BFS (breadth-first search) or DFS (depth-first search) with each state evaluated by a classifier (via a prompt) or majority vote.\nTask decomposition can be done (1) by LLM with simple prompting like "Steps for XYZ.\\n1.", "What are the subgoals for achieving XYZ?", (2) by using task-specific instructions; e.g. "Write a story outline." for writing a novel, or (3) with human inputs.\nAnother quite distinct approach, LLM+P (Liu et al. 2023), involves relying on an external classical planner to do long-horizon planning. This approach utilizes the Planning Domain Definition Language (PDDL) as an intermediate interface to describe the planning problem. In this process, LLM (1) translates the problem into “Problem PDDL”, then (2) requests a classical planner to generate a PDDL plan based on an existing “Domain PDDL”, and finally (3) translates the PDDL plan back into natural language. Essentially, the planning step is outsourced to an external tool, assuming the availability of domain-specific PDDL and a suitable planner which is common in certain robotic setups but not in many other domains.\nSelf-Reflection#', metadata={'source': 'https://lilianweng.github.io/posts/2023-06-23-agent/'})

Wie wir sehen können, ist das Dokument bereits in Stücke aufgeteilt. Und der Inhalt der Daten ist über den KI-Agenten.

RAG-Kette mit Milvus-Vektorspeicher aufbauen

Wir werden einen Milvus-Vektorspeicher mit den Dokumenten initialisieren, der die Dokumente in den Milvus-Vektorspeicher lädt und einen Index unter der Haube aufbaut.

from langchain_milvus import Milvus, Zilliz
from langchain_openai import OpenAIEmbeddings

embeddings = OpenAIEmbeddings()

vectorstore = Milvus.from_documents(  # or Zilliz.from_documents
    documents=docs,
    embedding=embeddings,
    connection_args={
        "uri": "./milvus_demo.db",
    },
    drop_old=True,  # Drop the old Milvus collection if it exists
)

Durchsuchen Sie die Dokumente im Milvus-Vektorspeicher anhand einer Testabfrage. Schauen wir uns das oberste Dokument 1 an.

query = "What is self-reflection of an AI Agent?"
vectorstore.similarity_search(query, k=1)

[Document(page_content='Self-Reflection#\nSelf-reflection is a vital aspect that allows autonomous agents to improve iteratively by refining past action decisions and correcting previous mistakes. It plays a crucial role in real-world tasks where trial and error are inevitable.\nReAct (Yao et al. 2023) integrates reasoning and acting within LLM by extending the action space to be a combination of task-specific discrete actions and the language space. The former enables LLM to interact with the environment (e.g. use Wikipedia search API), while the latter prompting LLM to generate reasoning traces in natural language.\nThe ReAct prompt template incorporates explicit steps for LLM to think, roughly formatted as:\nThought: ...\nAction: ...\nObservation: ...\n... (Repeated many times)', metadata={'source': 'https://lilianweng.github.io/posts/2023-06-23-agent/', 'pk': 449281835035555859})]

from langchain_core.runnables import RunnablePassthrough
from langchain_core.prompts import PromptTemplate
from langchain_core.output_parsers import StrOutputParser
from langchain_openai import ChatOpenAI

# Initialize the OpenAI language model for response generation
llm = ChatOpenAI(model_name="gpt-3.5-turbo", temperature=0)

# Define the prompt template for generating AI responses
PROMPT_TEMPLATE = """
Human: You are an AI assistant, and provides answers to questions by using fact based and statistical information when possible.
Use the following pieces of information to provide a concise answer to the question enclosed in <question> tags.
If you don't know the answer, just say that you don't know, don't try to make up an answer.
<context>
{context}
</context>

<question>
{question}
</question>

The response should be specific and use statistics or numbers when possible.

Assistant:"""

# Create a PromptTemplate instance with the defined template and input variables
prompt = PromptTemplate(
    template=PROMPT_TEMPLATE, input_variables=["context", "question"]
)
# Convert the vector store to a retriever
retriever = vectorstore.as_retriever()


# Define a function to format the retrieved documents
def format_docs(docs):
    return "\n\n".join(doc.page_content for doc in docs)

Verwenden Sie die LCEL (LangChain Expression Language), um eine RAG-Kette zu erstellen.

# Define the RAG (Retrieval-Augmented Generation) chain for AI response generation
rag_chain = (
    {"context": retriever | format_docs, "question": RunnablePassthrough()}
    | prompt
    | llm
    | StrOutputParser()
)

# rag_chain.get_graph().print_ascii()

# Invoke the RAG chain with a specific question and retrieve the response
res = rag_chain.invoke(query)
res

"Self-reflection of an AI agent involves the process of synthesizing memories into higher-level inferences over time to guide the agent's future behavior. It serves as a mechanism to create higher-level summaries of past events. One approach to self-reflection involves prompting the language model with the 100 most recent observations and asking it to generate the 3 most salient high-level questions based on those observations. This process helps the AI agent optimize believability in the current moment and over time."

Herzlichen Glückwunsch! Sie haben eine grundlegende RAG-Kette auf der Grundlage von Milvus und LangChain erstellt.

Filtern von Metadaten

Wir können die Milvus Scalar Filtering Rules verwenden, um die Dokumente anhand von Metadaten zu filtern. Wir haben die Dokumente aus zwei verschiedenen Quellen geladen, und wir können die Dokumente nach den Metadaten filtern source.

vectorstore.similarity_search(
    "What is CoT?",
    k=1,
    expr="source == 'https://lilianweng.github.io/posts/2023-06-23-agent/'",
)

# The same as:
# vectorstore.as_retriever(search_kwargs=dict(
#     k=1,
#     expr="source == 'https://lilianweng.github.io/posts/2023-06-23-agent/'",
# )).invoke("What is CoT?")

[Document(page_content='Fig. 1. Overview of a LLM-powered autonomous agent system.\nComponent One: Planning#\nA complicated task usually involves many steps. An agent needs to know what they are and plan ahead.\nTask Decomposition#\nChain of thought (CoT; Wei et al. 2022) has become a standard prompting technique for enhancing model performance on complex tasks. The model is instructed to “think step by step” to utilize more test-time computation to decompose hard tasks into smaller and simpler steps. CoT transforms big tasks into multiple manageable tasks and shed lights into an interpretation of the model’s thinking process.\nTree of Thoughts (Yao et al. 2023) extends CoT by exploring multiple reasoning possibilities at each step. It first decomposes the problem into multiple thought steps and generates multiple thoughts per step, creating a tree structure. The search process can be BFS (breadth-first search) or DFS (depth-first search) with each state evaluated by a classifier (via a prompt) or majority vote.\nTask decomposition can be done (1) by LLM with simple prompting like "Steps for XYZ.\\n1.", "What are the subgoals for achieving XYZ?", (2) by using task-specific instructions; e.g. "Write a story outline." for writing a novel, or (3) with human inputs.\nAnother quite distinct approach, LLM+P (Liu et al. 2023), involves relying on an external classical planner to do long-horizon planning. This approach utilizes the Planning Domain Definition Language (PDDL) as an intermediate interface to describe the planning problem. In this process, LLM (1) translates the problem into “Problem PDDL”, then (2) requests a classical planner to generate a PDDL plan based on an existing “Domain PDDL”, and finally (3) translates the PDDL plan back into natural language. Essentially, the planning step is outsourced to an external tool, assuming the availability of domain-specific PDDL and a suitable planner which is common in certain robotic setups but not in many other domains.\nSelf-Reflection#', metadata={'source': 'https://lilianweng.github.io/posts/2023-06-23-agent/', 'pk': 449281835035555858})]

Wenn wir die Suchparameter dynamisch ändern wollen, ohne die Kette neu zu erstellen, können wir die Laufzeitketteninterna konfigurieren. Definieren wir einen neuen Retriever mit dieser dynamischen Konfiguration und verwenden wir ihn, um eine neue RAG-Kette zu erstellen.

from langchain_core.runnables import ConfigurableField

# Define a new retriever with a configurable field for search_kwargs
retriever2 = vectorstore.as_retriever().configurable_fields(
    search_kwargs=ConfigurableField(
        id="retriever_search_kwargs",
    )
)

# Invoke the retriever with a specific search_kwargs which filter the documents by source
retriever2.with_config(
    configurable={
        "retriever_search_kwargs": dict(
            expr="source == 'https://lilianweng.github.io/posts/2023-06-23-agent/'",
            k=1,
        )
    }
).invoke(query)

[Document(page_content='Self-Reflection#\nSelf-reflection is a vital aspect that allows autonomous agents to improve iteratively by refining past action decisions and correcting previous mistakes. It plays a crucial role in real-world tasks where trial and error are inevitable.\nReAct (Yao et al. 2023) integrates reasoning and acting within LLM by extending the action space to be a combination of task-specific discrete actions and the language space. The former enables LLM to interact with the environment (e.g. use Wikipedia search API), while the latter prompting LLM to generate reasoning traces in natural language.\nThe ReAct prompt template incorporates explicit steps for LLM to think, roughly formatted as:\nThought: ...\nAction: ...\nObservation: ...\n... (Repeated many times)', metadata={'source': 'https://lilianweng.github.io/posts/2023-06-23-agent/', 'pk': 449281835035555859})]

# Define a new RAG chain with this dynamically configurable retriever
rag_chain2 = (
    {"context": retriever2 | format_docs, "question": RunnablePassthrough()}
    | prompt
    | llm
    | StrOutputParser()
)

Probieren wir diese dynamisch konfigurierbare RAG-Kette mit verschiedenen Filterbedingungen aus.

# Invoke this RAG chain with a specific question and config
rag_chain2.with_config(
    configurable={
        "retriever_search_kwargs": dict(
            expr="source == 'https://lilianweng.github.io/posts/2023-06-23-agent/'",
        )
    }
).invoke(query)

"Self-reflection of an AI agent involves the process of synthesizing memories into higher-level inferences over time to guide the agent's future behavior. It serves as a mechanism to create higher-level summaries of past events. One approach to self-reflection involves prompting the language model with the 100 most recent observations and asking it to generate the 3 most salient high-level questions based on those observations. This process helps the AI agent optimize believability in the current moment and over time."

Wenn wir die Suchbedingung ändern, um die Dokumente nach der zweiten Quelle zu filtern, da der Inhalt dieser Blogquelle nichts mit der Frage zu tun hat, erhalten wir eine Antwort ohne relevante Informationen.

rag_chain2.with_config(
    configurable={
        "retriever_search_kwargs": dict(
            expr="source == 'https://lilianweng.github.io/posts/2023-03-15-prompt-engineering/'",
        )
    }
).invoke(query)

"I'm sorry, but based on the provided context, there is no specific information or statistical data available regarding the self-reflection of an AI agent."

---

Dieses Tutorial konzentriert sich auf die grundlegende Nutzung der Milvus LangChain-Integration und einen einfachen RAG-Ansatz. Für fortgeschrittene RAG-Techniken lesen Sie bitte das Advanced RAG Bootcamp.