Qwen Me Anime (Experimental Checkpoint)Qwen Me Anime is a specialized anime-centric checkpoint trained using the Musubi Trainer tool. This model is designed to capture the vibrant aesthetics of modern anime while maintaining a flexible structure for various character designs.⚙️ Training Details<ul><li>Anime Dataset: 2,500 curated anime images.</li><li>Regularization Dataset: 3,000 images to maintain model stability.</li><li>Status: Experimental. As a work-in-progress, you may encounter occasional imperfections or inconsistencies in certain generations.</li></ul>💡 Key Characteristics<ul><li>Anime Stylization: Deeply tuned for sharp lineart and cel-shaded coloring.</li><li>Versatility: Capable of handling a wide range of character prompts and backgrounds.</li><li>Optimized for Qwen: Built to leverage the unique architecture of the Qwen vision-language foundation.</li></ul>🚀 Recommended Settings (Important!)To achieve the best results and overcome the current "experimental" limitations, it is highly recommended to use this checkpoint in combination with a lighting LoRA:<ul><li>Essential LoRA: Qwen Image Lighting</li><li>LoRA Versions: Both 4-steps or 8-steps versions work excellently.</li><li>Why? This helps stabilize the output, improve contrast, and clean up visual artifacts.</li></ul>📝 Usage Suggestions<ul><li>Prompting: Focus on descriptive anime tags (e.g., masterpiece, high quality, vibrant colors, detailed background).</li><li>Sampler: Euler or DPM++ series (Experiment based on your preference).</li><li>CFG Scale: 1 - 3 (Depending on your specific LoRA setup).</li></ul>⚠️ DisclaimerThis model is still in its beta/experimental phase. Feedback and sample images from the community are greatly appreciated to help improve future versions!

Files with SHA256: e19d6d0bb0e00332388048e5bd10109e643ce988ed678cfbedd74371b2b5b237