<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/0f6b6abd-75c6-4a33-be22-2497f0efc937/width=525/0f6b6abd-75c6-4a33-be22-2497f0efc937.jpeg" /><h3 id="ai">本人郑重声明：本模型禁止用于训练基于明星、公众人物肖像的风格模型训练，因为这会带来争议，对AI社区的发展造成不良的负面影响。</h3><h3 id="heading-31">本模型注明：训练素材中不包含任何真人素材。</h3>国风偏写实版本推荐-Recommended Realistic Version of National Style：<a target="_blank" rel="ugc" href="https://civarchive.com/models/77650/guofengrealmix">https://civarchive.com/models/77650/guofengrealmix</a>基于SDXL1.0训练的国风4-National Style 4 Based on SDXL1.0 Training:<a target="_blank" rel="ugc" href="https://civarchive.com/models/118009/4-guofeng4">https://civarchive.com/models/118009/4-guofeng4</a>一些答疑TIP：国风2或国风3以及大版本下的小版本，都是独立存在，不同版本的画风，关键词，对应内容都会有相应区别，建议下载时可以根据效果图或者模型的独立介绍去下载。TIP2：国风2代与3代的区别：我对大版本的划分是训练质量不同(比如素材分辨率，训练设置所消耗的显存量，有时候也会和素材量有关)。===========<h2 id="update-history">更新历史 - Update History</h2>2023.6.29 增加了v3.4版本2023.4.05 增加了v3.3版本2023.3.02 增加了v3.2_light版本2023.2.27 增加了safetensors格式-2023.6.29 added v3.4 version2023.4.05 added v3.3 version2023.3.02 added v3.2_Light version2023.2.27 added safetensors format===========<h2 id="model-introduction">模型介绍 - Model Introduction</h2>欢迎使用GuoFeng3模型 - 这是一个中国华丽古风风格模型，也可以说是一个古风游戏角色模型，具有2.5D的质感。GuoFeng3:原始模型GuoFeng3.1:对GuoFeng3人像进行了微调修复GuoFeng3.2:如果你不知道选择GuoFeng3还是GuoFeng2，可以直接使用此版本GuoFeng3.2_light:通过GuoFeng3.2融合了基于 Noise Offset 训练的Lora使得模型能够画出更漂亮的光影效果(Lora:epi_noiseoffset/Theovercomer8's Contrast Fix)GuoFeng3.3:此版本是基于3.2的一次较大的更新与改进，可以适配full body，即使你的tag不太好，模型也会对画面进行自动修改，不过因此模型出的脸会比较雷同。GuoFeng3.4:此版本重新进行了新的训练，适配全身图，同时内容上与前几个版本有较大不同。并调整了整体画风，降低了过拟合程度，使其能使用更多的lora对画面与内容进行调整。其它个人模型：同款模型画风的Lora版本：<a target="_blank" rel="ugc" href="https://civarchive.com/models/11352/guofeng3lora">https://civarchive.com/models/11352/guofeng3lora</a>西幻模型：<a target="_blank" rel="ugc" href="https://civarchive.com/models/10818/westmagic">https://civarchive.com/models/10818/westmagic</a>-Welcome to GuoFeng3 model - This is a Chinese gorgeous antique style model, which can also be said to be an antique game role model with a 2.5D texture.GuoFeng3: original modelGuoFeng3.1: The portrait of GuoFeng3 has been fine-tuned and repairedGuoFeng3.2: If you don't know whether to choose GuoFeng3 or GuoFeng2, you can use this version directlyGuoFeng3.2_Light: Through GuoFeng3.2, Lora based on Noise Offset training is integrated to enable the model to draw more beautiful light and shadow effects (Lora: epi_noiseoffset/Theovercolor8's Contrast Fix)GuoFeng3.3:This version is a major update and improvement based on 3.2, which can adapt to full bodies. Even if your tag is not very good, the model will automatically modify the screen, but therefore the faces produced by the model will be quite similar.GuoFeng3.4: This version has undergone new training to adapt to the full body image, and the content is significantly different from previous versions.At the same time, the overall painting style has been adjusted, reducing the degree of overfitting, allowing it to use more Lora to adjust the screen and content.Lora version of the same model painting style:<a target="_blank" rel="ugc" href="https://civarchive.com/models/11352/guofeng3lora">https://civarchive.com/models/11352/guofeng3lora</a>Western fantasy model:<a target="_blank" rel="ugc" href="https://civarchive.com/models/10818/westmagic">https://civarchive.com/models/10818/westmagic</a>===========<h2 id="3lora-guofeng3dlclora">国风3可选Lora - GuoFeng3DLC(Lora)</h2>1.梦 Dream (<a target="_blank" rel="ugc" href="https://civarchive.com/models/44310/dream-based-on-guofeng3">https://civarchive.com/models/44310/dream-based-on-guofeng3</a>)2.武墨 WuMo (<a target="_blank" rel="ugc" href="https://civarchive.com/models/47728/wumo">https://civarchive.com/models/47728/wumo</a>)===========<h2 id="recommended-settings">建议的设置 - Recommended settings</h2>Sampling steps:30 or 50Sampler:DPM++ SDE KarrasVAE:vae-ft-mse-840000 <a target="_blank" rel="ugc" href="https://huggingface.co/stabilityai/sd-vae-ft-mse-original/tree/main">stabilityai/sd-vae-ft-mse-original at main (</a><a target="_blank" rel="ugc" href="http://huggingface.co">huggingface.co</a>)======如果你的出图全身图时出现脸部崩坏建议删除full body关键词或者使用脸部自动修复插件：国外源地址：<a target="_blank" rel="ugc" href="https://github.com/ototadana/sd-face-editor.git">https://github.com/ototadana/sd-face-editor.git</a>国内加速地址：<a target="_blank" rel="ugc" href="https://jihulab.com/xiaolxl_pub/sd-face-editor.git">https://jihulab.com/xiaolxl_pub/sd-face-editor.git</a>-If you experience facial collapse during the full body image, it is recommended to delete the full body keyword or use the facial automatic repair plugin:Foreign source address: <a target="_blank" rel="ugc" href="https://github.com/ototadana/sd-face-editor.git">https://github.com/ototadana/sd-face-editor.git</a>Domestic acceleration address: <a target="_blank" rel="ugc" href="https://jihulab.com/xiaolxl_pub/sd-face-editor.git">https://jihulab.com/xiaolxl_pub/sd-face-editor.git</a>=====关键词 - key word:<pre><code>best quality, masterpiece, highres, 1girl,china dress,Beautiful face</code></pre>负面词 - Negative words:<pre><code>NSFW, lowres,bad anatomy,bad hands, text, error, missing fingers,extra digit, fewer digits, cropped, worstquality, low quality, normal quality,jpegartifacts,signature, watermark, username,blurry,bad feet</code></pre>更好的负面词 - Better negative words:<pre><code>(((simple background))),monochrome ,lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry, lowres, bad anatomy, bad hands, text, error, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry, ugly,pregnant,vore,duplicate,morbid,mut ilated,tran nsexual, hermaphrodite,long neck,mutated hands,poorly drawn hands,poorly drawn face,mutation,deformed,blurry,bad anatomy,bad proportions,malformed limbs,extra limbs,cloned face,disfigured,gross proportions, (((missing arms))),(((missing legs))), (((extra arms))),(((extra legs))),pubic hair, plump,bad legs,error legs,username,blurry,bad feet</code></pre>如果想元素更丰富，可以添加下方关键词 - If you want to enrich the elements, you can add the following keywords<pre><code>Beautiful face,
hair ornament, solo,looking at viewer,smile,closed mouth,lips
china dress,dress,hair ornament, necklace, jewelry, long hair, earrings, chinese clothes,
architecture,east asian architecture,building,outdoors,rooftop,city,cityscape</code></pre>

国风3 GuoFeng3

通过融合了基于 Noise Offset 训练的Lora使得模型能够画出更漂亮的光影效果(Lora:epi_noiseoffset/Theovercomer8's Contrast Fix)预览图的对比偏暗是特别加的效果tag，并非模型本身的风格，只是表现模型能够适应更真实的光影效果。我个人觉得右边的效果比较像照相的时候不开闪光灯，光线更加真实。============By integrating Lora based on Noise Offset training, the model can draw more beautiful light and shadow effects (Lora: epi_noiseoffset/Theovermeter 8's Contrast Fix)The contrast of the preview image is dark, which is a special effect tag. It is not the style of the model itself, but the performance model can adapt to a more realistic light and shadow effect. Personally, I think the effect on the right side is more like taking pictures without flash, and the light is more realistic.

Files with SHA256: f42edd43aa30455ce5c3d68ba7cb284b2516e9285a9e49c9813a83544c875e6b