<h2 id="this-is-a-lora-for-inducing-a-cartooncomic-feel-loosely-inspired-by-the-styles-of-unidentifiedtf-and-kannelart.-it-does-not-aim-to-replicate-said-styles.-atn2y1d5t">This is a LoRA for inducing a cartoon/comic feel, loosely inspired by the styles of <a target="_blank" rel="ugc" href="https://twitter.com/UnidentifiedTF">UnidentifiedTF</a> and <a target="_blank" rel="ugc" href="https://twitter.com/kannelArt">KannelArt</a>. It does not aim to replicate said styles.</h2>It is strongly influenced by style of the used checkpoint, so it works best with anime/cartoon models.No keyword/trigger. Meant to be used within 0.8-1.2 weight range.Goes crazy above 1.4, desaturates below 0. In negatives, v1 gives a painting/drawing vibe, v2 leans more into a stylized photo feel.<h3 id="adetailerhires.fix-recommended!-5rltgdpc3">ADetailer/HiRes.fix recommended!</h3>Strengths:<ul><li>girls (and feminine/anime boys)</li><li>portraits and closeups</li><li>stylized hair and eyes</li><li>saturated colours (stylistically, not like a VAE)</li><li>tight clothes</li><li>sfw and light nsfw</li></ul>Weaknesses:<ul><li>closely following pose prompts (ControlNet helps greatly)</li><li>some hairstyles are impossible due to stylization</li><li>covered shoulders</li><li>loose clothes</li><li>nudity and hardcore nsfw</li></ul>

KTF Style

Second versionThis version is less likely to produce random polygons in the background and it should be more consistent across different checkpoints, but is not compatible with as many models as v1.It is not a direct improvement over v1, rather v2 aims for a softer, rounder, more comic-like style compared to it. It also moves its attention to the whole picture, instead of just focusing on mainly the face and hair. As such, v2 inherits most of the strengths and weaknesses of v1.It does, however, hate covered shoulders even more than v1.

Files with SHA256: 687ada3b7d8fdf1ca1d486915313706aa23f09cf50a1db7ed1438bb898e906d6