This lora is trained for 4000 steps, it will generate flat vector illustration.

it requires upscaling.

It is style lora and trained on 14 images.

cannot change style of people more than 8 in a single image.

works perfect for upto 4-6 people.

This is the caption used while training lora

"change style of image to detailed flat vector illustration, while maintaining same composition, face details and face structure. flat colors, 4k resolution, ultra sharp, subtle gradient, solid plain color background."

IF YOU TRAINED LORA SIMILAR TO THIS PLEASE SHARE.