Advance RAG with Azure & LlamaIndex

Поделиться
HTML-код
  • Опубликовано: 6 сен 2024
  • With new collaboration between Azure AI Search and LlamaIndex, enabling developers to build better applications with advanced retrieval-augmented generation (RAG) using a comprehensive RAG framework and state-of-the-art retrieval system.
    ☎️ Do you need any career or technical help? Book a call with me: calendly.com/m...
    Code used in the video are under reference section in Discord Channel:
    Discord Channel: / discord
    ******************
    LET'S CONNECT!
    *******************
    Join Discord Channel: / discord
    ✅ You can contact me at:
    LinkedIn: / mohammad-ghodratigohar
    Email: mo.ghodrati95@gmail.com
    Twitter: / mg_cafe01
    🔔 Subscribe for more cloud computing, data, and AI analytics videos
    by clicking on the subscribe button so you don't miss anything.
    #RAG #Azure #llamaindex #Chat_with_data #Advance_Rag

Комментарии • 3

  • @Shoaibkhan-oj3oe
    @Shoaibkhan-oj3oe 4 месяца назад +1

    I have a csv file with text data, the data set has 3 columns. One is the main title, then the sub title and third point related to the title. I want to set a QnA bot on it. Basically the bot should be able to handle any queries regarding the dataset whether it is a summary of the points mentioned in it or a complete retreival of all the points. The answers should always be detailed and also gpt should use its own knowledge to enhance on based of the retreived data. I tried RAG but rag just provides me information on the chunks retreived but not able to give summarisation or is not able to give all the points from the sub topic. What is the best way to implement such QnA on such dataset. It only has text data. Please any help would be appreciated. I am really stuck on this. I tried asking this in your discord channel as well

  • @CaribouDataScience
    @CaribouDataScience 4 месяца назад +2

    Hey, where can I get one of those jumpers?

    • @MG_cafe
      @MG_cafe  4 месяца назад

      I will gift you one :) Send me an email