Fine-tuning

  • Resolved deonsa

    (@deonsa)


    3 years, 7 months ago

    Hi Meow,

    Could you perhaps give more detail on how to fine-tune the models to be used in this plugin? Do I need to install Python on our web server? I know there is a tutorial on OpenAI on how to do this, I’m just not exactly sure where to start.

    I would greatly appreciate a tutorial on this.

    Regards,

    Deon

Viewing 3 replies - 1 through 3 (of 3 total)
  • Plugin Author Jordy Meow

    (@tigroumeow)

    3 years, 7 months ago

    Hi @deonsa,

    I am going to add some tools to do this directly in the plugin 🙂 With a nice UI, and you’ll even be able to edit/modify the datasets you want the AI to train on.

    I’ll try my best to deliver this feature as soon as Monday (that will take the full weekend and Monday I believe)!

    You should check this: https://beta.openai.com/docs/guides/fine-tuning/preparing-your-dataset. The plugin will make this very easy, but it’s important to understand the overall process, and what kind of dataset is required.

    • This reply was modified 3 years, 7 months ago by Jordy Meow.
    Thread Starter deonsa

    (@deonsa)

    3 years, 7 months ago

    Hy Jordy,

    Thanks a lot for the fantastic support. I gave you a 5-star rating, would make it 6 if I could.

    Regards,

    Deon

    Plugin Author Jordy Meow

    (@tigroumeow)

    3 years, 7 months ago

    Hi @deonsa,

    It’s released 🙂 I made it super easy to upload dataset and train new AI models. However, it’s difficult to make a really good dataset, and even trickier to use it.

    I recommend you to try simple datasets, and then play with it in the playground (the AI Engine one or the Open AI one). When you start getting interesting results, we can think of how to handle the chabot cases (it’s trickier than it seems but we can do it, and we will).

Viewing 3 replies - 1 through 3 (of 3 total)

The topic ‘Fine-tuning’ is closed to new replies.