<h1 id="heading-140">Read the description before using, please!!!</h1>You can also use this and other resources in <a rel="ugc" href="https://tensor.art/u/693186869780861138/models">TensorArt</a>.<h3 id="heading-141">PonyXL</h3>Recommended strength 0.5 - 0.8*My recommendation: 0.7Trigger word:<pre><code>lanterby</code></pre>Base appearance (for custom outfits):<pre><code>blonde hair, short hair, bangs, x hair ornament, eyebrows, blue eyes,</code></pre>Official outfit:<pre><code>blue cardigan, off shoulder, long sleeves, open clothes, striped clothes, shirt, white shirt, collared shirt, cross bow tie, white skirt, miniskirt, pleated skirt, brown footwear, loafers,</code></pre>Recommendations to save a little time:<ul><li>*You can increase the weight of LoRA, but by doing so, you’ll sacrifice a bit of versatility (plus you may start to run into problems with the quality of eye generation in certain situations, such as shooting from behind), in exchange for a more anime style.</li><li>This is a bit basic but I'll mention it anyway: Take into consideration whether or not you'll use the "source_anime" and "source_pony" tags.</li></ul>All example images about this version in this post were generated using <a rel="ugc" href="https://civarchive.com/models/288584?modelVersionId=324524">AutismMix SDXL - confetti</a> model. (VAE in “Automatic”).

Files with SHA256: 92574e5b934e6c18e6f75b35932707f944d21a3d89676555a660aeaab5f8cf47