Warning about the Krea2 version. I just found out that it tends to make a lot of things longer, not just the hair. I will look into this in a v2. But until i make a v2 i have adjusted the recommended stable range to a range that should be ok. Outside of "Stable"-range use at your own risk XD
Krea2:
Stable from -2.5 to 2.5 (both Krea2 and ZIT)
ZIT:
Stable from -4 to 4
But works outside this range as well.
Works very well in combination with my Hair Volume Slider
All my work is 100% free. The only thing that really matters to me is people leaving a review, good or bad. I will NEVER remove or block anyone for leaving a bad review, so don't be afraid to give me criticism.
I'm stating this because when a creator blocks someone for leaving a bad review, that person can no longer see any of the creator's models, essentially muting their critics. This is something I'm highly critical of, and I believe it's a fundamental quality problem on this site.
Description
FAQ
Comments (7)
please, please, please, please, please, please, please, please, please, please, please, please, please, please, please, create a zoom in/zoom out and light on/off sliders for Krea2 !!!!
Its impossible to make a good slider for it as well. Just remember to prompt the feet an the head in the prompt and you should get a full body image.
Loraholic himself just published an article on how to achieve the desired outcome without any lora whatsoever. I found it very insightful and well written. Maybe give it a read. You can find it on his profile. Cheers :)
So, making a woman's hair short doesn't really give them the pixie cut I'd expect, but more of a boys face with a buzz cut. If the short hair pair could have more women with pixie cuts, that'd be nice.
A slider dont use a dataset in the way a normal LoRA does, so i can just add a pixie cut as a option. And a slider often drag in other elements when its pushed to far. So if you get a guy-face you should try to lower the value.
A slider should also be used in combination with prompts for best results. So try this and see if you still have the problem.
- turn down the strength
- Prompt what you want
- Use Slider to help push it the last stretch if prompts is not enough.
Do you use ai-trainer or musubi for slider training? Also, how many text or image pairs usually go into one of your sliders? Do you use an anchor prompt?
@phazei I have a pretty detailed guide in one of the comments on my new detail slider. That will answer most of your question. The dataset i use is just 10 random images from Krea2.