Type | |
Stats | 242 1,098 110 |
Reviews | (41) |
Published | Feb 15, 2024 |
Base Model | |
Training | Steps: 1,500 Epochs: 10 |
Usage Tips | Clip Skip: 1 Strength: 0.9 |
Trigger Words | SelectiveColorStyle 1colorpop 2colorpop |
Hash | AutoV2 AE4DC5ED93 |
About:
This lora is aimed at assisting with selective color prompts, especially ones with graphic novel-style inspirations. I have found that SDXL can be very hit and miss with understanding the concept of limiting colors without going into a completely monochromatic palette.
Helpful words and tags to add with the trigger:
Don’t forget to indicate what color(s) you want to appear! The lora left on it’s own may pick a color for you, but for some prompts it may default to a monochrome palette anyway without this input
1colorpop or 2colorpop (depending on how many colors you want to emphasize)
Selective color _____ (blue/pink/red/yellow/etc)
A genre “mood” such as “sci-fi mood” or “fantasy mood” (also dystopian, dark fairytale, horror, etc)
Recommendations (Models, Settings, Observations, etc.):
To be updated as I test!
So far, it has done fine with Extra Realistic XL and Niji style model I have tested in addition to SDXL.
I have noticed some of that dreaded overexposure when combined with lots of other loras, so please keep that in mind.
The lora was trained on clip skip 1, but has so far responded ok for me in the 2-4 range as well (since I know some people prefer those for their anime and 2.5 style prompts)
Please have fun playing with the lora if you try it out! I am open to suggestions and feedback. I am still learning, so constructive criticism and recommendations from this community are also always invited.