This mostly came from an experiment with a new model, but most of the images were generated with Anylora. some of my Niji model's

triggering words are "barbaragordon" "1 girl," "solo," "blue shirt," "short sleeves." recommend weight 0.7 Resolution 768x768, like ı said this was expriment,I didn't train with many images. I wasn't even planning on uploading. But it seems to work mostly alright.

Edit; Here are some things I discovered when using this Lora for best generated images, set the weight between 0.6-0.8, use a resolution of 912x912, and then upscale since the training images were of mediocre quality.Adding "short hair,red hair, brown eyes, belt" can help keep the design more consistent when reducing weight.