<h1 id="krea2-v1.0">Krea2 V1.0</h1>There we go. The Krea2 version also works great. And it respects the proportions of the background face so much better than other models. Strength of 0.6 is fine.I'd advise to not use artyclaw-bginset as trigger. It tends to create clawy hands :( But it works fine without.Have fun!<h1 id="anima-v1.0">Anima V1.0</h1>The first Anima version is out of the oven, and it works great!0.8 strength is a solid starting point. Two options: eyes and face - see triggers.Have fun ❤️<h1 id="illu-v2.0">Illu V2.0</h1>I re-trained the concept into two separate versions:V2.eyes and the more general V2.face. See the respective versions for base prompts.Start with 0.8 strength.Prompting for conflicting states (eg. "breasts out, covered nipples") works ok with the "face" version - see pinned examples below.<h1 id="illu-v1.0">Illu V1.0</h1>Creates a background zoom inset, based on <a target="_blank" rel="ugc" href="https://civarchive.com/models/2126928/p5-cut-in">p5 cut-in</a> but hopefully more general.Main trigger: bginset, cut-in, zoom layerSub triggers: {bgeyes|bgface|bgpussy|bgass}pussy and ass are less reliable. Use something like:<ul><li>bgpussy, pussy focus, pussy</li><li>bgass, ass focus, anus slip</li></ul>Example:<pre><code>bginset, bgeyes, eye focus, cut-in, zoom layer, {cowboy shot|feet out of frame|full body}, {low angle|dutch angle|straight-on|high angle}, 1girl, solo, solo focus, fighting stance, dynamic pose in foreground</code></pre>Strength variance is very high, approx. 0.5 to 1.5. Have fun and show off :)

Background Inset (Zoom Layer, Cut-In)

Trained on the same dataset as Anima.

artyclaw-bginset-krea2-v10.safetensors

Files with SHA256: c14203782767861f4db6b398219874347f7490a4176b39cd953127ad94699524