>>69449624NTA but Nice, welcome!
The model making stuff was ancient even before XL, we haven't really kept those up to date.
That being said model mixing isn't that hard just pick two models you like and use the checkpoint merger to merge them at specific ratio. Downside to doing with animagine you may lose character knowledge if you go too low ratio with animagine.
The LoRA making stuff should work for XL and yes XL learns better but is pretty resource intensive on the actual training. You can do on 8gb VRAM but I think you need to use specific settings for it.
