All embeddings were joined into one

  • I entered 46 Questions and Answers in FAQ mode by using your extension.

    Nevertheless, I have 1 vector at Pinecone. When I ask questions, my GPT tells “This model’s maximum context length is 4097 tokens, however you requested 7201 tokens (5701 in your prompt; 1500 for the completion). Please reduce your prompt; or completion length.”

    So, all 46 q/a joined into one long vector that overflows the GPT prompt.

Viewing 1 replies (of 1 total)
  • Plugin Author senols

    (@senols)

    Hi @porton

    Thanks a lot for reporting this. I am working on a new functionality to split large contents into smaller chunks. Will release this soon.

    As a workaround, you can add your questions in parts or you can use gpt-3.5-turbo-16k to get large context window.

    Hope this helps.

Viewing 1 replies (of 1 total)
  • The topic ‘All embeddings were joined into one’ is closed to new replies.