<h1 id="-5ressdxi9">交通工具设计-二维渲染模型（中性风格）</h1><h1 id="automotive-design-2d-rendering-lora-model-(characterized-by-a-neutral-style)-e4bpsdrlz">Automotive Design 2D Rendering LoRa Model (characterized by a neutral style)</h1>可以用于草图转草效图的工作流程中；It can be used in the workflow of sketch to 2D rendering conversion;<ul><li>渲染风格：偏中性的渲染风格，画面个人风格更少，相比起另外一个版本光影更加柔和，明暗交界线不锐利，透视更加正常，画面黑白对比度不会特别明显，色阶过度更加自然rendering style: A more neutral rendering style, with less personal style in the image. Compared with the other version, the light and shadow are softer, the boundary between light and dark is not sharp, the perspective is more normal, the black-and-white contrast of the picture is not very sharp, and the transition of color levels is more natural.</li><li>底模：SDXL 1.0 原版的就行 我用的是青龙大佬修复后的这个：<a target="_blank" rel="ugc" href="https://civarchive.com/models/117188/sdxlfixedvaefp16remove-watermark">sdxlFixedvaeFp16Remove_baseFxiedVaeV2Fp16.safetensors [1fa5725f4f]</a><a target="_blank" rel="ugc" href="https://civarchive.com/models/117188/sdxlfixedvaefp16remove-watermark">下载链接：https://civarchive.com/models/117188/sdxlfixedvaefp16remove-watermark</a>Base Model: The original version of SDXL 1.0 . I'm using the version fixed by the bdsqlsz:<a target="_blank" rel="ugc" href="https://civarchive.com/models/117188/sdxlfixedvaefp16remove-watermark">sdxlFixedvaeFp16Remove_baseFxiedVaeV2Fp16.safetensors [1fa5725f4f]</a>Download Link: <a target="_blank" rel="ugc" href="https://civarchive.com/models/117188/sdxlfixedvaefp16remove-watermark">https://civarchive.com/models/117188/sdxlfixedvaefp16remove-watermark</a></li></ul><ul><li>另外，我高度推荐使用 <a target="_blank" rel="ugc" href="https://github.com/lllyasviel/Fooocus">lllyasviel </a>Fooocus 来进行出图，可能控制严格程度有所下降，但是整体氛围效果更佳协调！Additionally, it's highly recommended to use lllyasviel Fooocus for rendering. While this might reduce the weight of control, the overall atmospheric effect is more harmonious.</li></ul><ul><li>受限于的SDXL1.0目前的 control-net 线控模型的质量，需要一些时间花在控制参数上的调整，如果需要较严格的线稿控制，我个人推荐 lineart_anime 预处理器，搭配 mistoLine_rank256 的 cn 模型，可以用参考图搭配ip-adapter风格引导来的到更直接的效果；Due to the limitations of the current control-net line control model in SDXL 1.0, some time needs to be spent on adjusting control parameters. If stricter line drawing control is needed, I personally recommend the mistoLine_rank256, paired with the cn model of lineart_anime. This combination allows for a more direct effect by using reference images alongside IP-adapter style guidance.</li><li>越低的控制度和越简洁干净的线稿使用效果越好，反之越严格、细节越多、模棱两可杂线越多效果越差，建议各位设计师使用前可以先收拾下自己的线稿来得到更好的效果；The lower the control weight and the cleaner the sketch, the better the effect. Conversely, higher control weight, more detailed, and more ambiguous sketch will yield worse results. It's recommended that designers tidy up their sketch before use to achieve better results.</li><li>具体的描述词见展示图，有些展示图是用fooocus出的，所以图片信息会特别多；触发词是car render style,For specific PROMPT, please refer to the display image, Some display images are produced by fooocus, so there will be a lot of info of the images. trigger word is "car render style."</li></ul>底下附上一些快照：Here are some snapshots for reference:<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/2c6a7551-3324-44e4-83ce-d18df36766e3/width=525/2c6a7551-3324-44e4-83ce-d18df36766e3.jpeg" /><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/afe43b2f-f6e6-4982-93e2-8bf0c452ebd6/width=525/afe43b2f-f6e6-4982-93e2-8bf0c452ebd6.jpeg" />两个版本对比：Comparison of two versions：<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/38293a41-8628-40a0-9cc8-0aba983b26b5/width=525/38293a41-8628-40a0-9cc8-0aba983b26b5.jpeg" /><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/bc9f2998-2c0c-403e-ae2d-37439da775c4/width=525/bc9f2998-2c0c-403e-ae2d-37439da775c4.jpeg" />

Files with SHA256: df87ff534c727c19ab25c152f87f1772f023e4c2a6f49f13dd2b78b0b6446f27