This project aims to recreate a actual Japanese woman named Ayumi. The 20,000 images in her photo album over the past 5 years were used as training data. This is an experiment to create LoRA that can learn everything without image processing and can respond to various scenes.
We welcome your feedback.