<h2 id="training-information">Training Information</h2>Trained in <a target="_blank" rel="ugc" href="https://github.com/ThetaCursed/Anima-TrainFlow">Anima-TrainFlow</a><ul><li>This LoRA is trained using Anima Base with the total image curated dataset of 170+ of Delga from Monster Ancient Cline/Mugen.</li></ul><h3 id="promptparameter-settings:">Prompt/Parameter Settings:</h3>Lora Strength: .7 - .9<h2 id="base">Base</h2>Activation Tag:<pre><code>1boy, delga, furry </code></pre>Reinforcement Tags:<pre><code>dragon, green eyes, blond hair, arm tattoo, tail, </code></pre>Negative Tags:<pre><code>black penis, purple background</code></pre><h2 id="outfit">Outfit</h2>Delga Outfit:<pre><code>clothes, vest, clothes around waist,white tank top, fingerless gloves, gauntlets, pants, </code></pre><h2></h2><h3 id="after-thoughts">After thoughts</h3><ul><li>This LoRA version was trained to a dataset that I loosely curated that leans in more a beefier and buff version of Delga, hence the name "XXL Delga".</li><li>When generating background, the dataset lacks variety within the background, so there will be moments the background will be purple.</li><li>Despite having a default style, it has the tendencies in generating black or latex like material on his genitals. Also, some occasion, generating a latex like clothes despite not being added within the prompt. So I highly recommended to add them within your negative prompts (or use NegPip which allows negative prompts within your main prompt) and specified a specific color for his genital like beige color.</li></ul>

Files with SHA256: 0a52468096a5c205a10038213091dc3196e7769685375591c1afb5c0dc737f22