This is my first attempt to train a LoRA model, and the character I chose is Lucy from Fairy Tail. When I first watched this anime ten years ago, Lucy's clear love and hate and cheerful personality still impressed me.
This time, the training results are released mainly as a record. The generated images have a strong sense of brushstrokes, which may be because the line drawing dataset I used accounts for a large proportion. As a novice who has just started model training, I am still learning and exploring, and I hope to generate more characters of different styles in the future. Welcome everyone to point out the shortcomings and make progress together!