Try size: 768 x 1024
Weights: 0.5~1.0
Clip skip: 2
Training material from AI generate pictures and bing
Pictures from kantanmix_v16LCMTurbo
Note: This lora is not necessary. The base model has already trained the "grilled corn" very well. It is based on this to enhance the diversity, make "grilled corn" look more delicious, the girl's expression more enjoyable, and the night market background more textured.