The VAE and text encoder are not baked in.
Use the following generation settings:
Positive Prompt: masterpiece, best quality, newest, year 2025, realistic
Negative Prompt: worst quality, low quality, blurry, jpeg artifacts, fewer digits, extra digits, bad hands, bad anatomy, watermark, 4koma
Sampler: DPM++ SDE Normal, 12 steps
Resolution: 832x1216 (Portrait) / 1216x832 (Landscape) / 1024x1024 (Square)
CFG Scale: 4
Hires: 2x-AnimeSharpV4, Upscaled by 1.5x, Denoise Strength 0.15, at least 5 Steps
Shift: 3.5
Using adetailer is strongly recommended. Anzhc's YOLO face model was used for this gallery. Models used in this merge are
Sciamano240 Style LoCon
jellytits-7 Style LoCon
NeoArtCore Style LoCon
Dandon Fuga Style LoCon
Zumidraws Style LoCon
Description
Initial release.