alright here's another random not so random mixture of various loras 3 checkpoint basis including the last and a few loras trained by me..

the profile picture was created with

euler 30 steps

8.5 cfg

latent 30 steps x 2

clip skip 2 is more varied but 1 works decent too or is it the other way around?

Best Quality, Outlined, Refined, Post-processed, Realistic, ultra-HD, (details:1.2), photo-realistic, girl, age 23, and specified clothes, it has good scenery if you include that word thanks to putting in my notaforest lora mixture,

negative: Grayscale, blurry, low quality worse quality, ugly, deformed, distorted, disfigured, mutated, bad hands, bad faces, extra limbs, bad ???????, anime, cartoon, painting,

it's kind of a close up ish picture taken heavy, it's hard to get it to do anything else. also unless you put ????? ???? in negative it will more than likely have ??????. even if you specify clothes....

also at time it can be very weird. one time you get blurred heavy background the next you get a bleeding real life epic scene.

I thought over all it's worth sharing. this was the 32nd blend in the chain in case you want to know lol.