How To Save OpenAI Embeddings In Database

Поделиться
HTML-код
  • Опубликовано: 28 окт 2024
  • НаукаНаука

Комментарии • 43

  • @mirekcolakov1730
    @mirekcolakov1730 Год назад +2

    I love this, thank you very much. It is just pure coding, no nonsense, just hard work. So many people are going the easy way and end up just talking and saying everything and nothing at the same time, riding the LLM wave. So, you Shweta, have my respect.

  • @gustavobenites5438
    @gustavobenites5438 Год назад +1

    That was exactly what I have been waiting for. Cant wait ur next vídeo showing How to query our own index.

  • @joaopsim
    @joaopsim Год назад

    Thank you, Shweta! Your explanation was very clear.

  • @AlgorithmicEchoes
    @AlgorithmicEchoes Год назад +4

    Please make a tutorial to create a streamlit front end for it. It'll be very helpful!

  • @Kalease54
    @Kalease54 Год назад

    Excellent. Your video are very easy to understand and informative. Very nice work

  • @Luan-i6z
    @Luan-i6z Год назад

    Thank you Shweta for great video. Can we use other document type such as pdf, images..etc?

  • @ziadullah8199
    @ziadullah8199 Год назад +1

    Hello Shweta, really nice video! After reading your article I tried this approach, however when i try to install `configuration` module it keeps throwing me error of module not found. I tried to pip install and various other approaches, also looked at stackoverflow, however couldn't figure out. It will be really nice if you could provide info about the python environment you are using. Your help would be appreciated in this regard.

    • @shweta-lodha
      @shweta-lodha  Год назад +4

      Thanks for watching Ziad. You need not to install configuration. This is just the py file, which I created to store my OpenAI key. In your case, you can remove configuration and directly provide your OpenAI key. Sorry for this confusion.

  • @bandui4021
    @bandui4021 Год назад +2

    Hi, thanks for the video. When coming to
    doc_store = Pinecone.from_texts([d.page_content for d in doc_texts], embeddings, index_name=index_name)
    it throws NameError: name 'Pinecone' is not define which is correct as we have not defined it.
    if i write pinecone.from.... it throws "AttributeError: module 'pinecone' has no attribute 'from_texts'"
    Where is my issue?
    thanks

    • @shweta-lodha
      @shweta-lodha  Год назад

      Looks like something has changed since I made this video. I would suggest you to look at the pinecone documentation and you will surely get some pointers.

    • @AlgorithmicEchoes
      @AlgorithmicEchoes Год назад +2

      check the spelling of 'pinecone' in from langchain,vectorstores import Pinecone. It should be capital P.

  • @IshaqKhan010
    @IshaqKhan010 Год назад

    can we rank the docs by question which has more accurate information? like we have CVs data and pass job description and return the best top CVs regarding JD via versa.

  • @yuvarajmani5907
    @yuvarajmani5907 Год назад

    This is awesome...

  • @jaymangukiya485
    @jaymangukiya485 Год назад +1

    Nice content but i am having one question,
    llm = OpenAI(......)
    How you are getting this class ?
    i think there is no class like it in openai sdk as well,
    Thank you in advanced.

    • @shweta-lodha
      @shweta-lodha  Год назад +1

      It is langchain

    • @xxtasee
      @xxtasee Год назад

      Great content indeed, I also had this issue i used "from langchain.llms import OpenAI" to resolve it

  • @IshmeetSinghahuja
    @IshmeetSinghahuja Год назад +1

    Thank you Shweta!! can you share your notebook? on github?

    • @shweta-lodha
      @shweta-lodha  Год назад

      Will this work medium.com/@shweta-lodha/give-permanent-memory-to-openai-embeddings-d4fdeb156fa3 ?

    • @IshmeetSinghahuja
      @IshmeetSinghahuja Год назад

      @@shweta-lodha perfect!! Thanks a lot 🙏

  • @abdullahhosam2717
    @abdullahhosam2717 Год назад

    I love it so much but question how we index those documents if I have multiple documents how I index retrieval them based on query.

    • @shweta-lodha
      @shweta-lodha  Год назад +1

      Glad that you like it. On indexing I have created few videos lately, please check out those. You will also get to know, how to save and query back your indexes

  • @xxtasee
    @xxtasee Год назад

    New suscriber, very good work!. Is there a way to run this code partially in Visual Studio Code (.py) and not Jupyter(.ipynb). If i do it in .py i have to run the whole code everytime and it duplicates the writing of the Vectors in PineCone. If i run it partially by putting #%% it doesn't work as it needs lines from the upper part. Best regards

    • @shweta-lodha
      @shweta-lodha  Год назад

      Please create functions and call them only when required

  • @elchinitowantan
    @elchinitowantan Год назад

    You are using python 2 or 3?, i can't use import configuration, this show me "ModuleNotFoundError: No module named 'configuration'" and i already install (pip3 install config = ok) and but this doesn't work :( help me

    • @elchinitowantan
      @elchinitowantan Год назад

      import configuration doesn't exist?

    • @shweta-lodha
      @shweta-lodha  Год назад +1

      Configuration is my own module, which I created to store API key. In your case, you can directly hardcode your key.

  • @elchinitowantan
    @elchinitowantan Год назад

    Can you help me with this error? PermissionError: [Errno 13] Permission denied: 'Store'

    • @shweta-lodha
      @shweta-lodha  Год назад

      Looking like you do not have permission to access store directory. Create in your local machine and then try

    • @elchinitowantan
      @elchinitowantan Год назад +1

      @@shweta-lodha I will do this suggestion

  • @ziadullah8199
    @ziadullah8199 Год назад

    Where I can find this code ?

    • @shweta-lodha
      @shweta-lodha  Год назад +1

      Here you go: medium.com/@shweta-lodha/give-permanent-memory-to-openai-embeddings-d4fdeb156fa3

  • @shardj5151
    @shardj5151 Год назад

    The video has a very high pitched continuous beep coming in and out throughout. It's painful to watch this.

    • @shweta-lodha
      @shweta-lodha  Год назад

      None report this and even I verified. Please try on some other device.

  • @shruthic4383
    @shruthic4383 Год назад

    I am getting the following error:
    raise MaxRetryError(_pool, url, error or ResponseError(cause))
    urllib3.exceptions.MaxRetryError: HTTPSConnectionPool(host='controller.us-west4-gcp-fre.e.pinecone.io', port=443): Max retries exceeded with url: /databases (Caused by NewConnectionError(': Failed to establish a new connection: [Errno 11001] getaddrinfo failed'))
    though I have logined in to Pinecone and asked for free storage . Can someone help me here