This model was trained on a dataset of 15 images. Some images were photoshopped to remove the crowd. The training went on for 10 epochs with 5 repeats each using Hassaku XL model (v1.3 - Style A). LoRA makes good 3D style, but may lose details in 2D.<h2 id="general-information:">General information:</h2><ul><li>Best weight: 0.7-0.9</li><li>Appearance: Sel1ne, blue eyes, long hair, blonde hair, lips</li><li>Outfit: circlet, red cloak, armor, pauldrons, breastplate, rerebrace, cuisses, armored boots, white gloves</li></ul>

Files with SHA256: ffa5238b45fb1cab230eb40b5d58c19a29f411f8b9f2935f27994f1b6df30c79