This is an updated version where the area around her eyes has been corrected and the pupils are also output more clearly.
Base model: Anything V5
Training environment: pixai.art LoRA learning
Training images: 512×768 40 images(Ver.1.0)/512×768 70 images(Ver.2.0)