A lora trained off of solid colors.
1.5 uses Dreamshaper v8, XL uses WowXLv3.
It works as a color booster, intensifying the color given in the prompt.
It has the opposite effect in negative, inhibiting the prompted color.
The XL version is more subtle, so you may need to use higher weights.
When used in the negative prompt, set to 2+ weight. This will eliminate the selected color from most of your generations.
Description
Trained on WOWXLV3
FAQ
Comments (3)
Also works with combinations of colors such as "cyan and magenta" or simply "blue black". I found it could go as high a a strength of 1 without completely losing the definition of an image. Occasionally produces a solid color, but not often, even at high strength.
That's because the Lora rank is set to 1. Something like a person or style may require 32 or more, but when the input data is basically a red pixel, you only need a tiny network to generate it! Coincidentally, the smaller a Lora is, the more flexible it is.
Awesome!! Thanks for XL version!!!
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.