I can't say it will be perfect on ZIT, but I believe it will do the job for now; Lora has only been trained with 43 images so far.
For now, it's a test version; don't expect anything elaborate.
A15A4S3X8RRJ4KNRWD9GPH8WH0.safetensors