RAGStack and DataStax Enterprise (DSE) 6.9 example

  1. Pull the latest dse-server Docker image and confirm the container is in a running state.

    docker pull datastax/dse-server:6.9.0-rc.2
    docker run -e DS_LICENSE=accept -p 9042:9042 -d datastax/dse-server:6.9.0-rc.2
  2. Install dependencies.

    pip install ragstack-ai-langchain python-dotenv langchainhub
  3. Create a .env file in the root directory of the project and add the following environment variables.

  4. Create a Python script to embed and generate the results of a query.

    import os
    from dotenv import load_dotenv
    import bs4
    from langchain import hub
    from langchain_openai import ChatOpenAI, OpenAIEmbeddings
    from langchain_community.document_loaders import WebBaseLoader
    from langchain_core.output_parsers import StrOutputParser
    from langchain_core.runnables import RunnablePassthrough
    from langchain_text_splitters import RecursiveCharacterTextSplitter
    import cassio
    from cassio.table import MetadataVectorCassandraTable
    from langchain_community.vectorstores import Cassandra
    # Load environment variables
    openai_api_key = os.getenv("OPENAI_API_KEY")
    # Initialize Cassandra
    cassio.init(contact_points=['localhost'], username='cassandra', password='cassandra')
        "create keyspace if not exists my_vector_keyspace with replication = {'class': 'SimpleStrategy', 'replication_factor': '1'};"
    # Create metadata Vector Cassandra Table
    mvct = MetadataVectorCassandraTable(table='my_vector_table', vector_dimension=1536, keyspace='my_vector_keyspace')
    # Web loader configuration
    loader = WebBaseLoader(
                class_=("post-content", "post-title", "post-header")
    docs = loader.load()
    # Document splitting
    text_splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=200)
    splits = text_splitter.split_documents(docs)
    # Vector store setup
    vectorstore = Cassandra.from_documents(documents=splits, embedding=OpenAIEmbeddings(), table_name='my_vector_table', keyspace='my_vector_keyspace', vector_dimension=1024)
    retriever = vectorstore.as_retriever()
    # Language model setup
    llm = ChatOpenAI()
    # Chain components
    def format_docs(docs):
        return "\n\n".join(doc.page_content for doc in docs)
    rag_chain = (
        {"context": retriever | format_docs, "question": RunnablePassthrough()}
        | hub.pull("rlm/rag-prompt")
        | llm
        | StrOutputParser()
    # Invocation
    result = rag_chain.invoke("What is Task Decomposition?")

    You should see output like this:

    Task decomposition involves breaking down a complex task into smaller and simpler steps to make it more manageable. Techniques like Chain of Thought and Tree of Thoughts help models decompose hard tasks and enhance performance by thinking step by step. This process allows for a better interpretation of the model's thinking process and can involve various methods such as simple prompting, task-specific instructions, or human inputs.

Was this helpful?

Give Feedback

How can we improve the documentation?

© 2024 DataStax | Privacy policy | Terms of use

Apache, Apache Cassandra, Cassandra, Apache Tomcat, Tomcat, Apache Lucene, Apache Solr, Apache Hadoop, Hadoop, Apache Pulsar, Pulsar, Apache Spark, Spark, Apache TinkerPop, TinkerPop, Apache Kafka and Kafka are either registered trademarks or trademarks of the Apache Software Foundation or its subsidiaries in Canada, the United States and/or other countries. Kubernetes is the registered trademark of the Linux Foundation.

General Inquiries: +1 (650) 389-6000, info@datastax.com