Press the follow button if you like my work. Also feel free to share your images in the gallery.
Trained with 40 images for PonyDiffusion. The samples were generated with waiANINSFWPONYXL_v130, Mistoon, AutismMix, ponyRealism_V22.
Trigger: freezer room
The lora leans more towards realism because of it's training data. Put "source_anime", "source_cartoon" in your prompt and "3d" in your negative prompt if you want a more cartoonish style.
Settings:
Lora weight: 0.8
Sampler: euler a, 28 steps
CFG: 4 - 7
Clip skip: 2