Hannah Owo V1
SD 1.5 Standard LORA
DownloadHannah Owo V1
dataset included cross eyed images. I didn't think about how that might be bad. Adjust prompt accordingly.
Examples use Realistic Vision V1.2, The Ally's Mix III: Revolutions and AbyssOrangeMix2 - SFW.
Play with weights. My examples use 0.9
I personally love The Ally's Mix III: Revolutions for these real loras. It makes very cute fanart of them.
Makeup is kind of overpowered. Might train again with additional images. Might need to have more images that focus on her face.