This LoRA is dedicated to capturing the appearance, style, and essence of Lamb, the graceful hunter and one half of Kindred from League of Legends.Unlike a pure style LoRA, this model was trained on both the UNet (visuals) and the Text Encoder (concepts). This allows it to understand the core attributes of Lamb as a character, providing greater consistency in generating her signature features, from her mask to her ethereal presence. It is well-suited for creating detailed character portraits, action scenes, and artistic illustrations.--- Usage Instructions ---Activation: To generate Lamb, use the primary trigger word <code>lamb (kindred)</code>. Reinforce the character's appearance with descriptive tags for best results.<ul><li>Primary Trigger: <code>lamb (kindred)</code></li><li>Essential Tags: <code>white hair</code>, <code>lamb mask</code>, <code>wolf mask</code>, <code>animal ears</code>, <code>hooves</code>, <code>glowing</code>, <code>holding bow</code></li><li>Negative Prompts: Consider adding <code>human ears</code> if you find the model generates both sets.</li></ul>Recommended Settings:<ul><li>LoRA Weight: <code>0.7</code> to <code>1.0</code>. A weight of <code>0.8</code> is a good starting point.</li><li>Model: Any SDXL-based model. Results will vary depending on the base model's style (e.g., anime, semi-realistic).</li><li>Sampling:<ul><li>Sampler: Euler Ancestral CFG ++</li><li>Steps: 20</li><li>CFG Scale: 1.4</li></ul></li></ul>--- Character &amp; Style Description ---This LoRA aims to reproduce Lamb's iconic design with high fidelity:<ul><li>Signature Masks: Reliably generates both the lamb mask she wears and the wolf mask that represents her partner, Wolf.</li><li>Ethereal Presence: Captures her otherworldly and ghostly nature, often complemented by a soft glow.</li><li>Core Anatomy: Consistently renders her white hair, lamb-like ears, and hooved feet.</li><li>Weapon of Choice: The LoRA is proficient at generating Lamb with her spirit bow in various poses.</li></ul>The overall artistic style is illustrative and clean, blending well with popular anime-style SDXL models to create vibrant and high-quality images of the Eternal Hunter.--- Training Details ---This LoRA was trained using a custom TOML configuration with the following key parameters:<ul><li>Base Model: NoobAI Vpred 1.0</li><li>Resolution: 1024x1024 (with bucketing enabled from 256 to 4096px)</li><li>Network Rank (Dimension): 64</li><li>Network Alpha: 32<ul><li>Note: A higher dimension than alpha was chosen intentionally. This scales down the weight adjustments, which can help in learning finer details and preventing the LoRA from becoming "over-baked."</li></ul></li><li>UNet &amp; Text Encoder Training: Both were trained (<code>network_train_unet_only = false</code>) for comprehensive character concept learning.</li><li>Optimizer: AdamW8bit</li><li>Learning Rate: <code>2e-4</code> (2×10−4)</li><li>LR Scheduler: <code>cosine_with_restarts</code> (3 cycles)</li><li>Advanced Noise Scheduling: <code>min_snr_gamma = 5</code> and <code>zero_terminal_snr = true</code> were used to improve image composition and training stability.</li><li>Precision: Full <code>fp16</code> for efficient training.</li><li>Batch Size: 1</li><li>Max Training Epochs: 30</li></ul>

Files with SHA256: 1bcd0f6a139d505aaf50eb0747f21fd6b2a40a9c8aa31c3d5ca42f8e79cbaab9