사용자 LoRA
wilhelm van astrea_locon
Trained on 80 images from the anime. Used frames of the anime to train it.
Credit goes to Cybermeow for the program used to gather the dataset
https://github.com/cyber-meow/anime_screenshot_pipeline
