The main character from the show. This model works well in all shapes and sizes. Compatible with other TIs and LoRAs. Her bow usually is polka dot, but will adjust for other descriptions. Expressions are not the best, so I recommend using an extra tool to adjust that. Clip skip is not required, but still helpful.

Training:

30 images (screen caps upscaled to 4k)

10k steps at a rate of .0005

Trained with NAI model