<h3 id="deltarune-loras-for-sdxl-(animagine-illustrious-pony)-and-sd-1.5">Deltarune LORAs for SDXL (Animagine, Illustrious, Pony) and SD 1.5</h3>SDXL:<ul><li>Trained on Animagine and Pony models</li><li>Covers 20 Deltarune characters with variations: Kris, Noelle, Susie (light/dark world versions), Jevil, Ralsei (Ch.1/2), Berdly (light/dark), Spade King, Lancer, Cyber Queen, Rouxls, Spamton (+Neo), Swatchling, Tasque Manager, Rudolph, Ant Tenna, Roaring Knight, Dess, Carol, Shadowguy, Pluey</li></ul>Prompting Tips:<ul><li>Place trigger words at beginning of prompt (except Pony model)</li><li>Pony model trained with shuffling, so you can use trigger words in different places with higher success</li><li>Additional details help (e.g., Kris's sweater type)</li><li>Caution with "masterpiece, best quality" in Animagine models, can cause gender bend and NSFW</li></ul>Legacy Models:<ul><li>Characters may struggle to interact unless similar</li><li>Can influence output style minimally</li></ul>Rework Information:<ul><li>See "about this version" section for specifics, same goes for other models</li></ul>Pony Model:<ul><li>Slightly undertrained, but sufficient for most characters</li></ul>SD1.5 Model:<ul><li>Lower quality than SDXL, requires more detailed prompting</li><li>Limited comparison data with other 1.5 models</li></ul>

Deltarune [XL, Illustrious, Pony, 1.5] LoRA

It's been trained using the Illustrious model, but in my tests models like NoobAI have had no problems, although v-pred models can be a bit messier than usual. I guess they aren't that far off from each other in terms of training.

Files with SHA256: 76827c1b771d0d236f98e9438655cf8fac48fca509a9193d09a924299dab7f8f