Features
It was traind with photographs of a legendary Japanese old school swimsuit, TYPE U-883.
Not imitaions, and a half of these are wetty pictures.
This model will be improved for your support. See Donations:
How to Use
Recommended LoRA weight is 0.2 to 0.4. (Or more at realistic models)
Watch out for your downloded file name. Write same name into prompts.
<lora:u883SchoolSwimsuit_v15:0.3>Negative Prompts: saunaWET and DRY mode
This model has different drawing effects, called "wet and dry".
Each has a different trigger word.
WET
<lora:u883SchoolSwimsuit_v15:0.3>,(wet u883 school swimsuit:1.1)Also use ”wet” tag together.
DRY
<lora:u883SchoolSwimsuit_v15:0.3>,(u883 school swimsuit:1.1)See also sample images.
Recommended SD models
Anything V3 or later
Realistic models also available
Training based model is ACertainty.
Term of Use
Not for corporate activities (Give me money)
Prohibit redistribution
Donate me if you feel, like that your life will be good
Thank you for 6 cup of coffee!(Mar.21,2023)
Coffee to me: https://ko-fi.com/kanaejun
BitCoin: 3H33ahQnpL84S4kSKWxCPzKMoEjtdgMBRy
LiteCoin: MWYA5K8jNo9AoCadvPpoMHs9bCEB23QYuk
Author
kanaejun, Japan
Description
Added training images
Changed training parameters. Dim up to 48, steps down to 7200.
Attempted layered-training
"down_lr_weight": "1,0,0,1,0,0,1,0,1,1,1,1", "mid_lr_weight": "0", "up_lr_weight": "1,1,1,1,1,1,1,1,0,1,1,0"