OverviewThis is a LoRA trained on Zatanna as she appears in the game Something Unlimited. It is carefully optimized to capture her distinct character design, signature outfit, and the game's recognizable 2D art style.❤️ Support the Original Artist!This LoRA was created out of deep admiration for the game's stellar art direction. Please support the original creator, GunsmokeGames! If you love using this model to generate images, consider backing their official Patreon to support their creative work and access their latest projects.<hr />🖼️ Preview Image Generation SettingsAll showcase images were generated directly using the parameters below. Stick to these settings to get the most accurate, cleanest representation of the character:<ul><li>Sampler / Method: <code>Euler a</code></li><li>Schedule Type: <code>Normal</code></li><li>Steps: <code>30</code></li><li>CFG Scale: <code>3.5 - 5</code></li></ul><hr />🚀 Usage Tips<ul><li>Trigger Word: <code>suzatanna</code></li><li>Recommended Weight: <code>0.7 - 1.0</code> (Adjust based on your checkpoint)</li></ul>📝 Prompting Guide<ul><li>Core Activation: Always include <code>suzatanna</code> at the absolute beginning of your prompt.</li><li>Outfit Tags: <code>heroine outfit, top hat, bowtie, tuxedo jacket, vest, white shirt, wrist cuffs, leotard, pantyhose</code></li><li>Character Accuracy (Eyes): To accurately recreate her signature look from the game, you may need to add <code>no pupils</code> to your positive prompt so she does not have pupils.</li><li>Outfit Control: If the costume is showing up in your images when you do not want it to, try putting <code>heroine outfit</code> in your negative prompt.</li><li>Anima Style Advice: Because Anima relies heavily on a hybrid of descriptive natural language and tags, explicitly describe the background and lighting to maximize composition quality.</li></ul><hr />📸 Share Your Creations!I would love to see what you create with this model! Please upload your generated art and images to the gallery below. Leaving a review and sharing your results helps me improve the model and shows others what it can do!

Files with SHA256: 42387e8256ae676bdc2fd7ac99a3494a6dcc1a19fc657ecec62209ed63a1b7cd