<h2>Update information</h2>V3.0 updated. Fine-tuned LoRA to improve the effects of generating characters with complex body limbs and backgrounds. The overall styling is more toward manga style rather than simple lineart. A lower CFG scale can make lines thinner.V3.0更新。对LoRA进行了微调，改善生成人物肢体和复杂背景的效果。整体更趋向漫画风而非纯线稿。低CFG scale可以让线条变细。<h2>Recommend settings</h2>Model: Anything V4.5VAE: Orangemix (the same with NAI)LoRA Strength: 1Sampler: DPM++ 2M KarrasSampling steps: 20CFG: 7Negative embedding: <a target="_blank" rel="ugc" href="https://civarchive.com/models/7808/easynegative">EasyNegative</a>、<a target="_blank" rel="ugc" href="https://civarchive.com/models/16993/badhandv4-animeillustdiffusion">badhandv4</a>Highres fix is also recommended.Note: all the LoRA names used in sample images are my local name, you need to change them to your saved LoRA filename!基模：Anything V4.5VAE：Orangemix（同NAI）LoRA强度：1采样器：DPM++ 2M Karras采样步数：20CFG：7负面embedding：<a target="_blank" rel="ugc" href="https://civarchive.com/models/7808/easynegative">EasyNegative</a>、<a target="_blank" rel="ugc" href="https://civarchive.com/models/16993/badhandv4-animeillustdiffusion">badhandv4</a>推荐使用高清修复。注意：样图中的LoRA名称是我本地文件名，要改成你保存的文件名！<h2>Difference</h2>Here's a comparision of V2.0 and V3.0 (locally called V3 and V4 on my machine - V2 on my machine is not published):V2.0和V3.0的对比（在我的本地机器上命名为V3和V4——本地的V2我并未发布）：<img src="https://imagecache.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/38354d83-7523-4334-c3de-1a3a493e5a00/width=525/38354d83-7523-4334-c3de-1a3a493e5a00" /><img src="https://imagecache.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/cc03ae1f-4391-4ff5-4850-d9478d78c600/width=525/cc03ae1f-4391-4ff5-4850-d9478d78c600" /><img src="https://imagecache.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/4c7eaa16-9f4d-4545-dc36-c2d3f43b9d00/width=525/4c7eaa16-9f4d-4545-dc36-c2d3f43b9d00" />Comparision of different CFG scale:不同CFG scale对比：<img src="https://imagecache.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/845cf24a-c5c6-4ee0-a999-d7130ecb7e00/width=525/845cf24a-c5c6-4ee0-a999-d7130ecb7e00" />

Anime Lineart / Manga-like (线稿/線画/マンガ風/漫画风) Style

Lineart style LoRA trained with offset noise, use <code>monochrome</code> or <code>lineart</code> to trigger, I highly recommend removing any color words from your prompt, like:<ul><li><s>purple eyes</s></li><li><s>blue</s> flower</li><li><s>red</s> dress</li></ul>线稿风LoRA，使用时请添加<code>monochrome</code> 或<code>lineart</code> 开启。推荐将prompt中与颜色相关的单词去掉。The demo images are generated w/ Anything V4.5 (orangemix VAE).如何利用ControlNet提取线稿（using ContorlNet to extract linearts from illustrations）：<a target="_blank" rel="ugc" href="https://www.bilibili.com/video/BV1Nv4y1a7ts">https://www.bilibili.com/video/BV1Nv4y1a7ts</a>

Files with SHA256: 9fb5ea0e8b421105296154510f86190da0540ea89fedddfd555b5fdb39a71c4e