How might I fine-tune a model in PremAI?

So I'm interested in the topic of self hosted LLM models, it's really appealing being able to have your own infrastructure (and pricing) but also being able to tailor the dataset the model will reference. I've been using ChatGPT for most of this year and I'm continuously feeding it random bits of text and articles so it has context of topics I want to query / compose content around.

I'm getting kind of tired of this process though! It's annoying trying to feed it enough text to let it better understand the tone for the persona / simulacra the model should respond as.

• How do you go about training a model if you have this text?

• Can someone build an interface that makes it easy to train OpenAI or Prem?