yeah hi im back from the dead apparently. soo, this one is actually made with the new ai toolkit thing, using a mere 21 images. the low count of images is definitely visible, but not that bad. which means, more putting necessary tags in negatives, yay..

activation: nerstyle

put those in negatives just in case: multiple views, 2girls, ghost, ghosts,

i screwed up with tho of the images in training, so the lora got fixated on tiny things, hence the low amount of images, mentioning again. once the artist makes makes more art, ill probably make a v2 of this