Another Lora trained with very few input images, so expect overfitting.

The sample prompts are also a bit unrefined - feel free to start removing tags that look extraneous or off.

Annoyingly, I had to redo all my sample images because of "lolita fashion, gothic lolita, lolita headband" even though these are all just clothing terms. Feel free to readd them to the prompt.

As always, it is trained NAI with full tags, which means you ought to use the tags in the sample images as your starting prompt, then refine them (you can remove a lot of tags).