Attention!, LoCon Version need install Extension: https://github.com/KohakuBlueleaf/a1111-sd-webui-locon
Update 2023-03-08
add LoCon-v1.0, small and better, only 2.5MB, recommond weight: 1.0
checkpoint: abyssorangemix3-aom3
no trigger words required
recommond weight: 0.8
you can use any prompts, even other characters
Description
50 images were used for training
network_dim = 1, network_alpha = 1
conv_dim = 1, conv_alpha = 1
output is 2.5MB.
FAQ
Comments (3)
Hey for the tag are you using the Dreambooth tokenizer? or just 1girl?
delete tags related to faces, heads, and hair