This is a test of 2D animation style lora trained on sd3. You can use full natural language to generate, control color, and add a variety of atmosphere to have better results. See the reference imageex.prompt:A woman with black hair, green eyes, leaking flowing light, in front of a black background, bust, she wears a suit, necklace, gold earrings, only eyes have colorA hooded man with black hair, head down, revealing golden swirling eyes in front of the black background, bust, one hand holding a luminous magic ball, scattering magic particles, gray and blue picture atmosphere, bringing a sense of mysteryblack light style, A man wearing sunglasses, yellow eyes, wearing a necklace, staring down, the whole picture is black, the face has a clear edge lightblack light style, Miku Hatsone uses a cool pose in the picture, a hand touches her face, her green eyes sparkle, and flash particles float in the air. The whole picture is dark, presenting a mysterious, red-purple atmosphereBlack lighting style, Miku Hatsone used a cool pose in the photo, a hand touching her face, her green eyes glinting, and flashing particles floating in the air. The green double ponytail hangs down, the whole painting is dark, showing a mysterious red and purple atmosphereBlack lighting style, a woman with white hair and golden eyes, wearing JK uniform, raised one hand to the screen, open hand, evil charm smile. In front of the green and dark background, the whole picture is colorless and filled with mysterious fognegativate_prompt:bad quality, poor quality, doll, disfigured, jpg, toy, bad anatomy, missing limbs, missing fingers,Noise, ripples1024*1536

SD3-First Anime Lora test

Higher training times and better performance in monochromatic scenes

Files with SHA256: d3c4312be1bf74cab9f2c0859c054e49a21c1b59a2f60fbd4ce4b952f1c83565