Uploaded image for project: 'Python Driver'
  1. Python Driver
  2. PYTHON-4241

Expose embedding dimension in openai-retrieval-plugin

XMLWordPrintableJSON

    • Type: Icon: Task Task
    • Resolution: Fixed
    • Priority: Icon: Unknown Unknown
    • 5.0
    • Affects Version/s: None
    • Component/s: AI/ML
    • None
    • Python Drivers
    • Not Needed
    • Hide

      1. What would you like to communicate to the user about this feature?
      2. Would you like the user to see examples of the syntax and/or executable code and its output?
      3. Which versions of the driver/connector does this apply to?

      Show
      1. What would you like to communicate to the user about this feature? 2. Would you like the user to see examples of the syntax and/or executable code and its output? 3. Which versions of the driver/connector does this apply to?

      Context

      Describe the background behind the problem.

      With OpenAI's latest text embedding models, the length of the vectors produced by the models is no longer fixed at 1536 as it was with their earlie 'ada' model. This caused confusion in their README. It also opens up a can of worms as we need to create our Index through the Atlas UI and numDimensions need to be set correctly. I have to document this properly, and it wasn't going to be esay.

      See Github Issue #426(https://github.com/openai/chatgpt-retrieval-plugin/issues/426) for detailed description.

      Definition of done

      What must be done to consider the task complete?

      Issue Accepted by maintainers.
      Pull-request completed.

            Assignee:
            casey.clements@mongodb.com Casey Clements
            Reporter:
            casey.clements@mongodb.com Casey Clements
            Votes:
            0 Vote for this issue
            Watchers:
            1 Start watching this issue

              Created:
              Updated:
              Resolved: