Base model: Anything V5Training environment: <a target="_blank" rel="ugc" href="http://pixai.art">pixai.art</a> LoRA learningTraining images: 512×768 70 images(v1.0/v2.0 both)★Adjust the LoRA strength to suit your environment.A lower strength is recommended, and in my environment around 0.5-0.6 or 0.7 was just right. If the intensity is too high, there is a high possibility that the eye area will become strange.I think v2.0 will be more stable if the LoRA strength is set lower. In my environment it was about 0.6.

Files with SHA256: afeeec74e7329758bea6beebd3eca709897b539cdc2f349c7adb0ad1341d2a15