from Hacker News

Winner of the SF Mistral AI Hackathon: Automated Test Driven Prompting

by carlcortright on 3/27/24, 5:31 PM with 15 comments

  • by bugglebeetle on 3/27/24, 6:04 PM

    Isn’t this just a very naive implementation of what DsPY does?

    https://github.com/stanfordnlp/dspy

    I don’t understand what is exceptional here.

  • by hitchstory on 3/27/24, 6:44 PM

  • by k__ on 3/27/24, 7:26 PM

    Just yesterday, I wrote an article about FT and learned about services like Entry Point AI.

    Seems like an awesome idea. I'm curious how long it will take to get a model on a reasonable level.

    Phind is pretty good and also the fastest model I used recently, so I'd assume it's quite small, no?

  • by carlcortright on 3/27/24, 6:06 PM

    Post Author: getting a lot of requests so scaling the backend. Standby.

  • by toisanji on 3/27/24, 7:06 PM

    can you please add more info on the page to show why it is important and how its helpful

  • by imranq on 3/27/24, 6:39 PM

    Shouldn't this be LoRA training?

  • by paradite on 3/27/24, 5:42 PM

    Yeah I've been thinking about this lately.

    LLMs come and go.

    Prompt engineering techniques come and go.

    But eval / labelled dataset is always useful once you built it.

  • by carlcortright on 3/27/24, 5:31 PM

    “Learning” by prompting, calculating the loss against evals, and updating the prompt