Miyuutsu's picture
Update README.md
02e1bf0 verified
---
license: other
license_name: faipl-1.0-sd
license_link: https://freedevproject.org/faipl-1.0-sd/
base_model:
- Miyuutsu/Kawaii_Kittopia_Catelier
- Minthy/RouWei-0.7
pipeline_tag: text-to-image
tags:
- merge
---
v2 has been through so many merges I don't even know anymore.
Best quality prompts: `masterpiece, best quality`
Optional additional quality prompts: `newest, absurdres, highres`
Negative prompts: `worst quality, low quality, watermark`
Optional additional negative prompts: `old, early, signature, text, bad quality, lowres, bad anatomy, bad hands, multiple views, abstract, japanese text, censored, sign, scan artifacts, jpeg artifacts, sketch, light particles, mutated hands`
This one isn't as picky about settings.
### Old description:
Versioning method: v{Merge_Method}.{Kittopia_Merge_Method}.{rouwei_Major_Version}.{rouwei_Sub_Version}-{Model_Iteration}
Quality Prompts: `masterpiece, best quality`
Negative Prompts: `worst quality, low quality, watermark`
Most prompts from both NoobAI and rouwei should work well. For artists try both `by {artist_name}` as well as just `{artist_name}`
Model is VPred ZSNR and has both metadata and tensors set correctly. Please ensure you are using a compatible UI.
Sampler: Euler
Scheduler: `Simple` (recommended), `Normal` or `SGM Uniform`
Steps: `30+`
CFG: `3~5`