v4 is an upgrade on v3, featuring on reduced meaningless random objects and color, better compositions, better eyes while maintaining most of the good features of v3 like nice color and light effects, obedience...
V4 has baked VAE (which is called V1.5prunedVAE or animevae in some other places) so you don't have to load extra VAEs
_
For better face details please enable ADetailer
Recommended parameters:
Base image resolution: 960 * 640 (3:2), vice versa
CFG Scale: 7
Clip Skip: 2
Highres Fix: use latent upscale (latent upscale method is bilinear, on comfyui you need to select it in Latent Upscale Node, but on webui it's just the default highres fix method just called "latent") to upscale the base image to 1536 * 1024 (1.6 times, 3:2, vice versa)
Upscale: use Ultimate SD Upscale with Controlnet Tile (My Prompt is More Important mode) enabled to upscale the highres fixed image to 2K, 4K or whatever
Update 2024/1/12
_
v4 is an upgrade on v3, featuring on reduced meaningless random objects and color, better compositions, better eyes while maintaining most of the good features of v3 like nice color and light effects, obedience...
V4 has baked VAE (which is called V1.5prunedVAE or animevae in some other places) so you don't have to load extra VAEs
_
For better face details please enable ADetailer
Recommended parameters:
Base image resolution: 960 * 640 (3:2), vice versa
CFG Scale: 7
Clip Skip: 2
Highres Fix: use latent upscale (latent upscale method is bilinear, on comfyui you need to select it in Latent Upscale Node, but on webui it's just the default highres fix method just called "latent") to upscale the base image to 1536 * 1024 (1.6 times, 3:2, vice versa)
Upscale: use Ultimate SD Upscale with Controlnet Tile (My Prompt is More Important mode) enabled to upscale the highres fixed image to 2K, 4K or whatever