A lora based on the character shirogane naoto from persona 4
Trained with 60 images
Image info
Model used anything v4.5
Notes
"grey eyes" helps as a large number of images were her shadow form(yellow eyes)
Training settings?
The most important part of training is getting a good dataset!
Training settings can be found by installing the additional networks extension in auto1111
Extensions -> available -> load from -> additional networks
Restart webui
place lora into stable-diffusion-webui\extensions\sd-webui-additional-networks\models\lora
To view the training settings
Go to additional network tab -> choose model -> training paramters
Description
FAQ
Comments (2)
Why does yours look cleaner than my shirogane model or does base anything v4.5 look worse than pruned version
Focusing on her dancing outfit prolly was a good idea
I tested out your shirogane and it seems to work just fine. Its most likely your prompts that are creating the slightly blurry effect, try and copy my prompts and replacing my lora with yours and it should work fine
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.