2026/03/15 Version 2.0 - Action &amp; Pose Expansion This is the major update for Gordon Quid. I have significantly expanded the dataset to include complex actions, vehicle interactions, and diverse emotional states.What's New in v2.0:<ul><li>Expanded Actions: Now supports dynamic poses like driving cars, riding motorcycles, bicycles, horses, and even piloting futuristic machines.</li><li>Emotional Depth: Improved consistency for expressive states, including crying, terrified, and lazy sleeping modes.</li><li>Multi-View Support: Added back, side, and front-view training data to stabilize Gordon's structure from any angle.</li><li>SDXL/Illustrious Optimization: Trained at 1024x1024 resolution with Rank 64/Alpha 32 for better lineart and color preservation.</li></ul>Recommendations:<ul><li>Trigger Word: gordon quid (Join as needed：western cartoon, 2d, flat color, thick lineart)</li><li>Recommended Weight: 1.0 - 1.6 (Depending on base model, It depends on the complexity of the actions and the scene.).</li><li>Tips: If you find the model trying to add "noses" in some base models, please use (no nose:1.5) or flat face in your positive prompt, and nose in your negative prompt.</li></ul>版本 2.0 - 動作與姿勢擴充版 這是 Gordon Quid LoRA 的重大更新版本。我大幅擴充了數據集，加入了豐富的動作、載具互動以及多樣的情緒狀態。v2.0 新增重點：<ul><li>動作擴充： 現在支援更多動態姿勢，例如駕駛汽車、騎乘機車、腳踏車、騎馬，甚至還有操控未來載具！</li><li>情緒細節： 優化了對各種表情的控制，包括大笑、流淚、驚恐與慵懶的趴睡模式。</li><li>多視角支援： 補充了背面、側面與正面的訓練數據，確保戈登在任何角度下都能穩定成像。</li><li>SDXL/Illustrious 優化： 採用 1024x1024 高解析度訓練，Rank 64/Alpha 32 配置，能更好還原美式卡通的粗線條與平塗質感。</li></ul>使用建議：<ul><li>觸發詞： gordon quid(視情況加入：western cartoon, 2d, flat color, thick lineart)</li><li>推薦權重： 1.0 - 1.6 (視底模而定跟動作及場景複雜度而定)。</li><li>小撇步： 如果在某些寫實底模上出現「長鼻子」的問題，請在正向提示詞加入 (no nose:1.5), flat face，並在負向提示詞加入 nose，即可完美修正。 ------------------------------------------------------------------------------------- About this LoRA: This is a LoRA of Gordon Quid from the animated series Catscratch. It is trained on the Illustrious XL (SDXL) architecture. It excels at capturing his exaggerated expressions, distinctive orange and white markings, and the sharp Western cartoon lineart style.</li></ul>Usage Tips:<ul><li>Trigger Word: gordon quid</li><li>Recommended Weight: 0.8 - 1.2 (Depending on the base model, 1.2 often gives the best results).</li><li>Recommended Prompt: gordon quid, western cartoon, 2d, flat color, thick lineart, solo</li><li>Base Model: Works best with Illustrious XL or other SDXL-based anime/cartoon models (like Pony Diffusion V6 XL).</li><li>Settings: Recommend CLIP Skip: 1 or 2.</li></ul>Have fun creating your own Gordon Quid adventures! 關於這個 LoRA： 這是一個來自動畫《貓爪抓》(Catscratch) 的 Gordon Quid 的 LoRA 模型，基於 Illustrious XL (SDXL) 架構訓練。它能完美捕捉他誇張的表情、標誌性的橘白相間毛色，以及粗線條的美式卡通畫風。使用建議：<ul><li>觸發詞 (Trigger Word)： gordon quid</li><li>推薦權重 (Weight)： 0.8 - 1.2 (根據使用的底模不同，建議設在 1.2 效果通常最好)。</li><li>推薦提示詞 (Prompt)： gordon quid, western cartoon, 2d, flat color, thick lineart, solo</li><li>適用底模： 建議搭配 Illustrious XL 或其他基於 SDXL 的動漫/卡通模型 (例如 Pony Diffusion V6 XL)。</li><li>設定： 建議 CLIP Skip 設定為 1 或 2。</li></ul>玩得開心，創造屬於你自己的 Gordon Quid 冒險吧！

Gordon_Quid_Catscratch

Version 1.0 - Initial Release This is the first version of the Gordon Quid LoRA trained on the Illustrious XL (SDXL) architecture.Training Details:<ul><li>Base Model: Illustrious XL v2.0</li><li>Dataset: 21 high-quality images capturing various expressions, poses, and props.</li><li>Training Specs: Trained using Kohya_ss with Rank 32, Alpha 16.</li></ul>Notes: This model captures his iconic "western cartoon" style, vibrant colors, and expressive face. Highly recommended for creating dynamic poses and funny scenarios. Feel free to share your results! 版本 1.0 - 首次發布 這是 Gordon Quid LoRA 的第一個版本，基於 Illustrious XL (SDXL) 架構訓練。訓練細節：<ul><li>基礎模型 (Base Model)： Illustrious XL v2.0</li><li>數據集： 21 張高品質圖片，涵蓋了各種表情、動作與道具。</li><li>訓練規格： 使用 Kohya_ss 進行訓練，設定為 Rank 32, Alpha 16。</li></ul>

Files with SHA256: 73844ace2c2402bfe3ee7d3585dc01060ebb15e0bcaeb6fa7aeaca935a103fce