This is an update on my previous Castlevania Style LoRA's. This is an image-trained LoRA for pony, this one has a larger dataset and should work in a1111/forge :)
checkpoint-e18_s504.safetensors