Verified: 2 years ago
SafeTensor
Type | |
Stats | 4,530 |
Reviews | (191) |
Published | Jul 13, 2023 |
Base Model | |
Hash | AutoV2 BBEB9014B8 |
Recommended Settings
VAE : kl-f8-anime2
Steps : More than 30 steps
CFG scale : 6-8
Sampler : DPM++ SDE Karras / DPM++ 2M Karras / DPM++ 2M SDE Karras
Hires upscaler : 4x-AnimeSharp / Latent (nearest-exact)
Negative prompts : (verybadimagenegative_v1.3:0.7), (worst quality, low quality:1.4), (multiple views:1.2), (variations:1.2), (monochrome), watermark, (elf ears), (lips)
All samples I uploaded were made with highres. There are more samples in my pixiv. https://www.pixiv.net/users/86784786
Model B is more realistic than Model A