Ginny Di, born June 9, 1990, is an American YouTuber, cosplayer, and DnD role player. The LoRA was trained on 84 publicly available photos in 1024x1024, 896x1152, 832x1216, and 768x1344. The training data did include some of her many, many unique and amazing cosplay costumes.Recommendations<ul><li>Use trigger ginnydi and LoRA weight 0.6 - 0.8</li><li>Photorealistic models such as <a target="_blank" rel="ugc" href="https://civarchive.com/models/43977/leosams-helloworld-sdxl-base-model">LEOSAM's HelloWorld</a> and <a target="_blank" rel="ugc" href="https://civarchive.com/models/133005/juggernaut-xl?modelVersionId=348913">Juggernaut XL</a> tend to produce the best results. These, however, may need a few tries due to the diversity of the dataset.</li><li>Trained with Clip Skip 1, results without Clip Skip vary.</li><li>Positive prompts turquoise hair, green and blonde hair, pink lipstick, smiling, and elf ears may be interesting. Bonus: Try cosplay, vamp and cosplay, elf!</li><li>You might need to specify hair, eyebrow, and eye colors.</li><li>Negative prompts, wrinkles and old may be helpful.</li><li>I didn't do much testing with CFG values, but 7 to 8 seems to be fine.</li></ul>

Files with SHA256: 408af43792643985419061357e8385d65c6c253766bd161f670d198293cc4dfd