Here's my first attempt on a Lora so it might have some issues. I still have lots to learn about training so any tips would be much appreciated. It's trained off of 60 1024x1024 images on Z Image Turbo using default parameters from Civitai's trainer.

prompt tips

2d, anime style,

1girl, reze, reze (chainsaw man), anime, purple hair, eyebrows hidden by hair, green eyes, hair between eyes, long bangs, medium hair, mature female, small breasts, black choker, choker, bare shoulders, black ribbon, collar, collared shirt, neck ribbon, ribbon, shirt, sleeveless, sleeveless shirt, white shirt, black shorts, shorts,

sampler euler a

steps 9

strength 0.8-0.9

(Sorry about the File name, you'll have to manually rename unless someone knows how I can fix that)