My first attempt at lora training, and it kinda sucks to be honest. It's difficult to get what you want and I'm not really sure what I did wrong. However, you can get some good results if you stick with it.

Parameters

If you do decide to use it, try 9:16 aspect ratio, i.e. 768x1344. As for the prompt, you will have to experiment, but here's what I found out:

A detailed illustration = more details, but colours are a bit muted

An illustration = less details, a little bit more colour

Digital art = Same details as "an illustration" but more vibrant.

Words like "Bold" or "Vibrant" will add some more colour into it.

Will there be updates?

I'll try to fix it, but the updates will depend on my luck. This is really hard and expensive/time consuming.