<h1 id="obligatory-disclaimer:-this-lora-is-intended-to-be-used-alongside-style-loras-andor-artist-style-tags-to-bring-out-its-full-potential.-all-example-images-use-wai-anima-and-the-&quot;haku89&quot;-artist-style-tag.">Obligatory Disclaimer: This LoRA is intended to be used alongside style LoRAs and/or artist style tags to bring out its full potential. All example images use WAI-Anima and the "haku89" artist style tag.</h1><h2 id="if-you-don't-see-all-4-example-images-make-sure-you're-logged-in-and-try-changing-browsing-level.">If you don't see all 4 example images, make sure you're logged in and try changing Browsing Level.</h2><h3 id="comments-and-gallery-posts-are-encouraged-and-appreciated-regardless-of-their-content!-(within-tos-of-course.)">Comments and gallery posts are encouraged and appreciated regardless of their content! (Within ToS, of course.)</h3>Start with a model strength of 1.0. If you find the model to be too stiff, try lowering it to 0.80.A character LoRA for the indie VTuber Erundel's Live2D model. Trained on a dataset of 40 images that consists largely of cut up screenshots, images generated/enhanced by ChatGPT, and a few fanarts.I've changed my LoRA training settings once more in hopes of better flexibility. I've swapped back to Anima Base since I think the generations look marginally better and background generation appears to be significantly nicer. I've kept the halved step count and Prodigy optimizer though. That being said, it's still not perfect and the are definitely still issues. Most notably, the bunny ears can incorrectly generate, not properly passing through her hat. Human ears can also sneak in when generating certain angles. Ultimately, not very glaring issues in practice but still something to keep in mind. The below and attached triggers are general suggestions. You can almost certainly get better and more consistent results by more specific prompting. I'm just lazy. Play around with it to see what works for you. Remember, Anima loves natural language so converting the listed triggers into natural language can result in more consistent generation. Add the "ahoge" trigger word if you're generating her without the hat.Detailers are highly recommended for the face and eyes.Manual inpainting or image refinement through ChatGPT are recommended if you're looking to generate small details.Probably won't be updated.All example images were generated without a refiner step. You can probably make your generations look much better with improved workflows.Triggers:ErundelL2D, short hair, blue hair, red eyes, large breasts, bunny ears, bunny tail,police hat, striped blue tie, blue shirt, black belt, dark blue pencil skirt, black tights, dark blue high heels,

Files with SHA256: c1b0ebca1b05bacc0f56e6b680d8f2eb15115af1106b558d2b3af971e54abe6b