Usage recommendationsSampling MethodDPM++ 2M Karras or DPM++ SDE Karrasafter several tests on Euler Ancestral, the face looks like childfor steps it's better at 20 - 40CFG scale 5 - 10 is bestResolution 512x768 (Portrait), 768x512 (Landscape)for VAE you don't need to use it because it's already bakedvaeIf the model is too bright, you can change it with another VAEPromptGirl : 1girl, female focusBoy : 1boy, male focusGrup : 3girl (use resolution 768x512 landscape for more effectiveness)Quality Prompt : masterpiece, best quality, incredibly absurdres, highres, high detail eyes, high detail backgroundNegative Prompt :<ul><li>(worst quality, low quality:1.4), EasyNegativeV2, ng_deepnegative_v1_75t, bad anatomy, interlocked fingers, text, signature</li><li>(worst quality, low quality:1.4), interlocked fingers, bad anatomy</li></ul>I don't recommend using bad-hands-5 or badhandv4 embeddings.better use Interlocked Fingers, Mutated hands, Bad handsHiresfix<ul><li>4x-AnimeSharp</li></ul>I recommend using 4x-AnimeSharp because the results are very goodset hires step 15 - 30<ul><li>R-ESRGAN 4x+ Anime6B</li></ul>set hires step 15 - 25<ul><li>Latent (nearest-exact)</li></ul>use latent (nearest-exact) with Euler Ancestralhires step you can set as before or 15 - 25if you use latent (nearest-exact), denoise strength can't be below 0.5 because the image will be broken (if you force it maybe 0.35 is the limit)if you copy image preview please first checking hiresfix setting and seedADetaileruse face_yolov8s.pt default settingif you feel the face doesn't match the style try reduce Detection model confidence threshold around 0.2 - 0.3, also you can reduce or increase Inpaint denoising strength if using LoRa or specific face of the characteryou can use ADetailer Prompt for example you want 2 different color eyes input "heterochromia, red eyes, blue eyes" or pupil shape, something like thatADetailer Negative Prompt functions like a regular negative prompt. you can leave it blankThose are all the settings that I used in the preview

Files with SHA256: 1cc8af4946908b1118b3894ed9d22e992d03b96e6354db608bd3bb0612b4ef0e