Fixed it so its no longer overbaked (Just like my cooking)
Works best at 0.8-1.0 strength
All preview images were made with kotosmix
Trigger word bianca
glasses are optional (type in red glasses)
Easynegative embedding https://huggingface.co/embed/EasyNegative/blob/main/EasyNegative.safetensors
Training settings
trained on 25 images
20 repeats per image
batch size 3
512x512
4 epoch
cosine with restarts 4
Standard learning rate(0.0001)
128 dim
64 alpha
default everything else
Description
Fixed it so its no longer overbaked (Just like my cooking)
Works best at 0.8-1.0 strength
Bianca3=Bianca2 Just a naming error on my part dont worry!
All preview images were made with kotosmix
Trigger word bianca
glasses are optional (type in red glasses)
Training settings
trained on 25 images
20 repeats per image
batch size 3
512x512
4 epoch
cosine with restarts 4
Standard learning rate(0.0001)
128 dim
64 alpha
default everything else
FAQ
Comments (3)
ty for the fix
What model do you use in training?
Just base novel ai
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.