
I'm going to redo this one but I thought I'd put it up if people want to mess with it. I wanted to get the pitch black eyes and void gradient limbs in XL so this is a first attempt. Trained with 100 Dall-E-3 images. It's pretty finicky with everything, I think I need to train with less image or something/
EDIT: Just added an 8 epoch that might work better for some people, also do NOT use adetailer for this, it basically kills all the details.
Added updated version - I would use 9 epoch version, but both can work. Should just be easier to get normalized looks, but it does look a little different from original. Some issues and still havn't got around to 1.5 as I have no models for it and what not.
描述:
Should be much easier to get what you want, there is some issues with background burn ins and just over training in general. Try to just use negatives or look at mine if you have issues. sd1.5 still on the back burner, working on something else atm.
训练词语:
名称: WB2.2.safetensors
大小 (KB): 223099
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success