test style with training images gereate with the Lilian lora
without lora: 
with lora:
compared with the Polly style has more soft colors, I'm new with style loras so i don't know if it's a good work or not 😅
Description
i had problems to train this style with the original dataset, so i've trained it with many less images, and seem quite good, at least with neo forge 😅