Large scale finetune of Illustrious with state of the art techniques and performance.
Dataset of 7M unique pictures (~2M with natural text captions) picked and balanced from 14M of anime art and other media, including private datasets. More detailed description on Civitai
Vpred version is now available!
Key advantages:
Better prompt following
Great aesthetic, anatomy, stability along with versatility
Vibrant colors and smooth gradients without trace of burning
Full brightness range even with epsilon
Knowledge of tens of thousands style and almost any character.
An addition, comparing with vanilla Illustrious and NoobAI:
No more annoying watermarks
No tags bleed and better prompt segmentation
No characters tags bleed and related side effects (unwanted outfits, style, composition changes)
Better coherence and anatomy
Artist styles look exactly as they should
Each style including base is stable without random fluctuations on different seeds
New knowledge
Dataset cut-off - 20th December 2024.
Features and prompting:
The model is designed to work both with short booru tag-based and long complex natural text prompts. The best result can be achieved using the combination of tags and some natural text phrases.
For tags classic danbooru-style comma-separated tags without underscores were used.
Basic settings:
~1 megapixel for txt2img, any AR with resolution multiple of 64 (1024x1024, 1152x, 1216x832,...). Euler_a, CFG 4..8 for epsilon/3..5 for vpred, 20..28steps. LCM/PCM/DMD untested, cfg++ samplers work fine. Highresfix: x1.5 latent + denoise 0.6 or any gan + denoise 0.3..0.55.
Please note that vpred version requires a lower CFG value.
There are only 4:
masterpiece, best quality
for positive and
low quality, worst quality
for negative
Nothing else. Meta tags like lowres have been removed, do not use them. Low resolution images have been either removed or upscaled and cleaned with DAT depending on their importance
Negative prompt:
worst quality, low quality, watermark
For best results keep it as clean as possible. Spamming of popular sequences will not improve results, since all related flaws have been solved, but will only lead to unwanted effects, biases and poor quality.
Artist styles:
The model knows over 35k of artist styles. List, grids with example on Mega. Used with by , will not work properly without it.
Use it in combination with booru tags, works great. Use only natural text after typing styles and quality tags. Use just booru tags and forget about it, it's all up to you. About 2M of pitures from dataset have hybrid natural-text captions made by Opus-Vision, GPT-4o, Gemini and ToriiGate
Version 0.7 comes with several improvements in prompt understanding and segmentation. For best performance keep track of CLIP 75 token chunks and how your prompt is separated into them.
Brightness/colors/contrast:
You can use extra meta tags to control it:
low brightness, high brightness, low saturation, high saturation, low gamma, high gamma, sharp colors, soft colors, hdr, sdr
Vpred version:
Vpred version of RouWei-0.7 is now available. It works flawlessly out of box without any burning or related issues. Consider to use lower CFG (3..5), other generation settings are same. Avoid to use some experimental samplers designed to reduce burning: nothing to fix here, it will only lead to low quality images.
Base model
Epsilon and vpred versions here have a brief aesthetic polishing after main training to improve small details and coherence. If you want to use RouWei in merges, extract something without bringing that last things, or finetune it - you can use base version of RouWei.
Model tends to generate NSFW images for corresponding prompts, consider to add extra filtering. Outputs may be inacurate and provocative and must not be used as a reference.
License:
Same as illustrious, please check out original page for limitation. Fell free to use in your merges, finetunes, ets. just please leave a link.
Thanks:
A number of anonymous persons, Bakariso, dga, Fi., ello, K., LOL2024, NeuroSenko, rred, Soviet Cat, Sv1., T. and other fellow brothers that helped.
