My Model Merge + Lora Stack Workflow that upscales to 4k under 12GB RAM

I'll get into more detail later when I can, but I've been asked to share this workflow and while I don't have the time to explain every node at the moment, I'm mainly just posting this as a way to see if I can get the workflow out into the wild. ComfyUI manager should make it possible to download all the nodes.

I spent a lot of time messing with AUTO CFG and its effects on CFG, samplers and LORA Stacking and Model Merging and I'm quite pleased with the result, the upscaling pipeline also has a full model merge and lora stack so it doubles as a refiner in order to get that extreme detail that I like.