Skip to content

Basic RAG

Here we show an example of building a basic RAG application with Indexify. We are going to upload content about Kevin Durant from Wikipedia and ask questions about KD's career.

Install the Indexify Extractor SDK, Indexify Langchain and the Indexify Client

pip install indexify-extractor-sdk indexify indexify-langchain

Start the Indexify Server

indexify server -d

Download an Embedding Extractor

On another terminal start the embedding extractor which we will use to index text from the Wikipedia page.

indexify-extractor download hub://embedding/minilm-l6
indexify-extractor join-server

docker run -d -v /tmp/indexify-blob-storage:/tmp/indexify-blob-storage -p 9500:9500 tensorlake/minilm-l6 join-server --coordinator-addr=host.docker.internal:8950 --ingestion-addr=host.docker.internal:8900 --advertise-addr=0.0.0.0:9500

Initialize Client

We will use the langchain wikipedia loader to download content from wikipedia and upload to Indexify. We will also use langchain to prompt OpenAI for the RAG application.

pip install --upgrade --quiet  wikipedia langchain_openai langchain-community

Instantiate the Indexify Client 

from indexify import IndexifyClient
client = IndexifyClient()

Create an Extraction Graph

extraction_graph_spec = """
name: 'wikipediaknowledgebase'
extraction_policies:
   - extractor: 'tensorlake/minilm-l6'
     name: 'wiki-embedding'
"""
extraction_graph = ExtractionGraph.from_yaml(extraction_graph_spec)
client.create_extraction_graph(extraction_graph)
# display indexes
print(client.indexes())

Add Docs

Now download some pages from Wikipedia and upload them to Indexify 

from langchain_community.document_loaders import WikipediaLoader
docs = WikipediaLoader(query="Kevin Durant", load_max_docs=10).load()


for doc in docs:
    client.add_documents("wikipediaknowledgebase", doc.page_content)

Perform RAG

Create a retriever to feed in data from Indexify. 

from indexify_langchain import IndexifyRetriever
params = {"name": "wikipediaknowledgebase.wiki-embedding.embedding", "top_k": 20}
retriever = IndexifyRetriever(client=client, params=params)

Initialize the Langchain Retriever, create a chain to prompt OpenAI with data retrieved from Indexify to create a simple Q and A bot 

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


template = """Answer the question based only on the following context:
{context}

Question: {question}
"""
prompt = ChatPromptTemplate.from_template(template)

model = ChatOpenAI()

chain = (
    {"context": retriever, "question": RunnablePassthrough()}
    | prompt
    | model
    | StrOutputParser()
)
Now ask any question about KD - 
chain.invoke("When and where did KD win his championships?")


'Kevin Durant won his championships with the Golden State Warriors in 2017 and 2018.'