IMPORTANT: to get the best results with this model and prevent oversaturation, you must use <a target="_blank" rel="ugc" href="https://arxiv.org/abs/2404.07724">"guidance limiter"</a> (included in <a target="_blank" rel="ugc" href="https://github.com/Panchovix/stable-diffusion-webui-reForge">reforge</a>, <a target="_blank" rel="ugc" href="https://github.com/pamparamm/ComfyUI-ppm">comfy node</a>) with it's default settings.information<a target="_blank" rel="ugc" href="https://civarchive.com/models/833294/noobai-xl-nai-xl">noobai vpred</a> merge, many models were used, but that is the base model as target.the only non-illustrious based model merged is <a target="_blank" rel="ugc" href="https://civarchive.com/models/573152">lustify (nsfw) </a>it's weight difference was added via perpendicular component a little to help character positioning/backgroundsmerge method<a target="_blank" rel="ugc" href="https://github.com/ljleb/sd-mecha">sd-mecha</a> via <a target="_blank" rel="ugc" href="https://github.com/ljleb/comfy-mecha">comfy-mecha</a> using many of it's methods like tensor sum, rotate, multiply quotient, geometric median etc. and a <a target="_blank" rel="ugc" href="https://files.catbox.moe/wwfvwv.py">modified karcher-mean script.</a> <a target="_blank" rel="ugc" href="https://github.com/win10ogod/Karcher-merge">(original source)</a>(the diff from the source is that all tensors are converted to a float type for the calculation and then it converts the final result back to the original data type to maintain model integrity.)recommended settingsinterface: <a target="_blank" rel="ugc" href="https://github.com/comfyanonymous/ComfyUI">comfyui</a> or <a target="_blank" rel="ugc" href="https://github.com/Panchovix/stable-diffusion-webui-reForge">reforge</a>samplers: res_multistep or euler (and their cfg++ variants)schedulers: either beta or ays+ (alignyoursteps but with sigma set at 0, <a target="_blank" rel="ugc" href="https://github.com/pamparamm/ComfyUI-ppm">it's available here)</a>step count: around 20cfg: 5 to 6prompts: include "very awa", any danbooru/gelbooru/e621 tags. (some natural language also works)negative prompt: low quality, worst quality, etc.

Files with SHA256: e8ae09da1314fe7e7ae1eb689350a2eb918f3c2ec65014373474b74cf4bac850