Today I built my first assistant for SDK in Java.
I’ve never built plugins before (thanks for the short but perfect documentation).
I’ve been working a lot with AI lately and haven’t built any more games… and haven’t been on the forum either. But today I would like to share a little bit.
Here’s a short video: https://youtu.be/5pp7Wh9V9vU
I have a GPU, so I like to use Ollama locally. Ollama has to run on your own computer, the codellama model has to be pulled. Of course you can also integrate other models.
I only spent a few hours on this small extension of SDK, so it’s just a showcase and can be expanded upon.
But every journey begins with the first step.
I would like to know from you, would you rather rely on GPT or do you find local models on your own GPU interesting too?
Thanks.
P.S. this first code is nowhere near as well thought out as this one: