本页面包含 MOSS 和 Anima 两个版本，请选择对应的使用说明。 This page includes both MOSS and Anima versions. Please follow the instructions for your model.<hr /><h1 id="recommended-settings">推荐参数/Recommended Settings</h1><h3 id="for-anima:">FOR ANIMA:</h3><ul><li>Trigger word:</li><li>V2 nor:@skp_n</li><li>V1 nor:@skpstyle</li><li>V1 dar:@skp_darkstyle</li><li>LoRA weight: 0.8~1</li><li>Sampler: er_sde + beta</li></ul><h3 id="for-moss">FOR MOSS：</h3><ul><li>no Trigger word！！！</li><li>LoRA weight: 1-1.3</li><li>Sampler: <span style="background-color:rgb(26, 27, 30);color:rgb(250, 82, 82);font-family:-apple-system, BlinkMacSystemFont, &quot;Segoe UI&quot;, Roboto, Helvetica, Arial, sans-serif, &quot;Apple Color Emoji&quot;, &quot;Segoe UI Emoji&quot;;font-size:16px">er_sde + beta OR euler + normal</li><li>cfg:3</li><li>steps:30-50</li></ul>____________________________________________________________________________________<h1 id="moss">MOSS</h1><h2 id="v2:">V2:</h2><ul><li>解决了不能使用质量提示词的问题，更换了配方和训练参数，整体效果更好。但注意质量词用不用，表现上仍有分别，原因不明，加上质量词，整体表现更规整、更细腻，不用会更有多笔触。我推测，在MOSS上，会将一些笔触、凌乱的线条当成是低质量的关系，所以一旦加上质量词，一些笔触会消失。</li><li>I've fixed the problem where quality prompt tags couldn't be used, and switched to a different recipe and training parameters — the overall results are better. Note, though, that whether the quality terms are used still makes a visible difference, for reasons I don't yet understand: with them the output is more orderly and more refined, without them it shows more brushwork. My hypothesis is that MOSS treats certain brushwork — loose, messy line work — as low quality, so once quality tags are added, some of that brushwork disappears.</li></ul><h3 id="v1.1">V1.1：</h3><ul><li>TARGET_RES:1536,更多步数训练。more steps.</li><li>使用 MOSS V1.1版本时，正向和负向提示词都不要加入任何质量词，否则可能无法准确还原画风。原因尚不明确。</li><li>When using the MOSS V1.1 version, do not add any quality tags to either the positive or negative prompt. They may prevent the LoRA from reproducing the art style accurately. The reason is currently unknown.</li></ul>_____________________________________________________________________________________________<h1 id="anima-normal-and-dark">Anima — Normal &amp; Dark</h1><h3 id="v2">V2：</h3><ul><li>新增了些许数据集。</li><li>TARGET_RES:1536，修改了配方和使用了新的训练器，整体表现更好，指令遵从度改善，光影、画面、阴影也有一定改善。</li><li>我也试着训练了2048的版本，但我认为过强，纹理过深。</li><li>added a small amount of new training data.</li><li>TARGET_RES:1536.Updated the training settings and switched to a new trainer, resulting in better overall performance, improved prompt adherence, and noticeable improvements in lighting, image quality, and shadows.</li><li>I also experimented with training a 2048-resolution version, but I personally find the effect too strong, with overly pronounced textures. </li></ul>同一位画师，两种画风。 Two styles from the same artist.<ul><li>**Normal：**明亮柔和，色彩清透。 Normal: Brighter and softer, with a lighter palette.</li><li>**Dark：**色彩浓郁，阴影厚重，光影对比更鲜明。 Dark: Richer colors, deeper shadows, and more dramatic lighting.</li></ul><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/6d567ae0-a3e4-4bfe-9bea-65bfb6ce32bf/original=true/6d567ae0-a3e4-4bfe-9bea-65bfb6ce32bf.jpeg" /><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/1ca9f783-f9c5-4c7b-972f-eb434994fdbc/original=true/1ca9f783-f9c5-4c7b-972f-eb434994fdbc.jpeg" />

[Anima/MOSS] sekali_p style

SKP_N_v2-2048.safetensors

SKP_N_v2.safetensors

anima v2-nor

Files with SHA256: 5b2d8fbe83e247deb25fccaae7b3ab2872dbbb8a6a8542d803a9b30dcca23b75