It's all coming together for Mike's seemingly "whacky" AI side projects, isn't it?
How exactly, if you don't mind expounding a bit, do you store the digital you over time? Is using local models a big part of that? What exact dataset are you building? Surely not trusting it to OAI on their cloud. And is it just a heap of various markdown files, or some sets of weights?
Login to reply
Replies (1)
I'm using an ever growing set of markdown files, to system prompt a local generic model. But I'm about to start training my own models based on those markdown files along with curated data feeds curated by more powerful cloud AI to create an actual "Mike model".
I'm doing deep research into training right now. This is a good starting point if you're interested:
Ever wondered how to use PyTorch to train an AI?
View quoted note →