<h1 id="heading-206"><span style="color:rgb(64, 192, 87)">竹羽系列</span>SDXL模型，发挥XL的特点，更清晰更敏锐，对TAG捕捉更精准，可以更好的发挥想象。</h1><p> </p><h2 id="heading-207">真实的模型：</h2><ul><li><p>动物皮肤细腻、细腻的人造</p></li><li><p>为了中国元素的捕捉更准确，中国人脸更习惯，但眼球部分容易出现蓝色眼球，需要增加TAG。</p></li><li><p>采样方式方面建议：Euler a、DMP++2M Karras以及3M系列</p></li><li><p>色调上：雨水度更高，画面会更鲜艳一些，如果想出做旧风格或者过滤效果的图，需要增加TAG或者LORA来完成。</p></li><li><p>标签：建议不要太复杂，简单明了更适合这个模型</p></li></ul><h3 id="heading-208"> 下面就是大家发挥想象探索的时间……</h3><p></p><p></p><p></p><h1 id="heading-209"><span style="color:#40c057">ZhuYu</span> series SDXL model showcases the characteristics of XL, making it clearer and sharper, and capturing TAG more accurately, allowing for better imagination.</h1><p></p><h2 id="heading-26">Realistic texture model:</h2><ul><li><p>The delicate and soft texture of animal hair</p></li><li><p>The capture of Chinese elements is also more friendly. The Chinese face has been improved a lot, but blue eyeballs are prone to appear in the eyeballs, and TAG needs to be added.</p></li><li><p>Suggestions for sampling methods: Euler a, DMP++2M Karras, and 3M series are all very friendly</p></li><li><p>In terms of color tone: with higher saturation, the image will appear more vibrant. If you want to create images with old styles or filter effects, you need to add TAG or LORA to complete them.</p></li><li><p>TAG suggestion: not too complex, simple and clear is more suitable for this model</p></li></ul><h3 id="heading-27">Here is the time for everyone to unleash their imagination and explore to the fullest……</h3>

Files with SHA256: d13dc50c023cd6a326562fa1e3160f8d788b387b50c33ee8883cb796ea14b149