Sakaki model for Pony Diffusion. Trained on 50 images over 30 epochs.
Weight of 1 works well on its own as well as with combining styles. The Sakaki tag seems to work fine on its own, and I have noticed no issues. For her outfits, try "serafuku" "school swimsuit" "gym clothes" and "yellow bikini".
Description
FAQ
Comments (2)
It does a good job of making her look like Sakaki, so it's quite the useful LORA. However, it likes to make her blurry. I'm guessing you probably trained it with screenshots from the Q-TEC upscale of Azumanga? If so, maybe don't train a LORA again using screenshots from a Q-TEC upscale.
I am curies about what base model the trainer used that don't have blurry eye detail with original animate style. Please tell me.