Description
Not perfect, but figured we could use some more chee here. Moved to working with Automatic1111. Still learning and using different settings and LORA. I just do these when I find some extra time. I have a bunch I want to img2img to fix some flaws. I'll try to post the good ones I have for now.
What I'm currently experimenting with:
Steps: 70, Sampler: DPM++ 3M SDE Exponential, CFG scale: 8, Seed: 2573677141, Size: 768x1024, Model hash: 1c39f05a56, Model: oneFORALLFurry_v20, VAE hash: e9ed949371, VAE: anythingKlF8Anime2VaeFtMse840000_klF8Anime2.safetensors, Clip skip: 2, TI hashes: "kkw-ph1: 644e5c86ce6d, EasyNegativeV2: 339cc9210f70, negative_hand-neg: 73b524a2da12, boring_e621_v4: f9b806505bc2, ng_deepnegative_v1_75t: 54e7e4826d53, kkw-ph1-neg: 185bc65f3975, updn: b4ae8ca1b247, ubbp: 047acf26d29c", Version: v1.7.0
DeNatural
MemberAbsolutely love it. What an incredible jump in quality.
Login to respond »