Note: This LoRa may only work in ComfyUI, or it may also work in other UIs which support all the same LoRa formats. It probably won't work as intended on the Civitai image generator. I don't know how to fix this, LoRa extractions just don't seem to be as compatible for some reason.This is a LoRa extracted from the weights difference given by the formula:<pre><code>Checkpoint A = Nova Furry v14 + uwustyle v2 * 0.8
Checkpoint B = Banana Splitz v3 + uwustyle v2 * 0.7 + Smooth Booster v3 * 0.25
Checkpoint C = Hassaku Illustrious v2 + uwustyle v3 * 0.7
Checkpoint D = Illustrious v0.1 + WAI Illustrious (unknown version) * 0.5 + Magic IL v2 * 0.4
Difference = (Checkpoint A + Checkpoint B * 0.5 + Checkpoint C * 0.4) - Checkpoint D</code></pre>Intended to be a sort of blend between uwustyle v2 and v3, plus a bit of furry and detail enhancement.The tags <code>cute</code>, <code>adorable</code>, and <code>chibi</code> are optional but help guide the style of the image.

uwumerge LoRa (Cute Furry | ComfyUI)

Initial version, still kinda experimenting with parameters but I liked the look of this version. Start with a strength of 1.0, or try 0.8 - 0.9 if halos appear around outlines, contrast goes up too much, or if anatomy gets weird.

Files with SHA256: af98ac1e13a23a58e21c3e399f1e0293e411336b2a0d05bc7fc1df417d1aa988