A LoRa for generating confused looking anime girls with interrogation marks on their heads.<a target="_blank" rel="ugc" href="https://civarchive.com/articles/18582/20-loras-and-images-the-credit-card-companies-dont-want-you-to-see">To see all my LoRAs, please (temporarily) disable X and XXX visibility, otherwise a lot will be hidden at the whims of credit card companies</a>. You can help support making models like this <a target="_blank" rel="ugc" href="https://civarchive.com/articles/9229/how-to-give-a-model-creator-50-buzz-at-no-extra-cost-possibly-negative-to-you">by posting stuff you've made with it to the gallery using the "add post" button</a>, that helps me earn back the training cost at no cost to you.V2:Trigger: C_L_A_G Pick one: ? or ?? Negative if using ?: ?? Pick one or both: confused, thinking Optional: | |, dot eyes, looking at viewer, head tiltGave tags a second pass (cleaning up a lot of garbage and consistently applying "?" versus "??" and camera so those should work better), cleared alpha channels, and removed a duplicate image. Was just going to make an Anima port till I saw how rough the original dataset was.V1:<a target="_blank" rel="ugc" href="https://civarchive.com/models/950882/confused-looking-anime-girls-with-interrogation-marks-on-their-heads">The PDXL version</a> won my poll on what to convert to ILXL, so it's my first LoRA of 2025. New for ILXL: “dot eyes” and “blank eyes” were added to applicable training data that didn’t have it.Tags: clagwimoth, confused, ?, portrait, Optional: | |, dot eyes, looking at viewer, blank eyesWas pretty solid from epoch 1, which is expected since the PDXL version worked that well and the concept is handled decently without a LoRA.

[IL] Confused looking anime girls with interrogation marks on their heads

Tag cleanup, removed alpha channels, removed duplicate image.

Files with SHA256: 9c0bfed749c3668e7c48e7af581fb96a470054a27ca67fe8bd0913f78227cc4c