Make sure to use UI that supports v-prediction, otherwise you will get a noisy blob!
Main parameters:
Steps: 28
CFG: 3-5
Sampler-scheduler: euler/euler a-normal, DDPM-exponential (better noses)
CFGZeroStar and sometimes Self-Attention Guidance: default for difficult backgrounds
Try to BREAK prompts into parts by 75 tokens max, this goes for most models..
Positive prompt:
number of characters (1girl/1boy), character name, artist prompt, scene/environment/camera angle, action, expression, items, quality prompts (masterpiece, best quality)some heavily stylized characters might require prompt: nose, realistic.. you can put nose into Adetailer
Negative prompt:
lowres, worst qualityeven longer negative should work, but I prefer simplicity.. tags worth noting include sketch, abstract, url
Enjoy..
Description
Doubled dataset, played with weight blocks. Slimes are not as translucent as in v2.
FAQ
Comments (5)
Dang that was a fast update from v2 to v3! Can't wait to try v3!
I let training run at night, so that helps. XD
@blinkdotleh ahhh that explains it :D
I was sold the moment I saw the image where there were pure black levels. That's what image models have been missing. Good contrast, it makes colors much better.
welcome to vpred, most vpred models do that ;)