秒变 AI 绘画大师,comfyui抓紧学起来
<h1 style="color: black; text-align: left; margin-bottom: 10px;">
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="//q9.itc.cn/images01/20240604/5879f525b86341468a17b10662cf237a.jpeg" style="width: 50%; margin-bottom: 20px;"></p><span style="color: black;">1、</span>SD主流 UI
</h1>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">Stable Diffusion(SD)<span style="color: black;">由于</span>其开源特性,有着较高的受欢迎程度,并且基于SD的开源社区及教程、插件等,都是所有<span style="color: black;">工具</span>里最多的。基于SD,有<span style="color: black;">区别</span>的操作界面,<span style="color: black;">能够</span>理解为一个<span style="color: black;">工具</span>的<span style="color: black;">区别</span>客户端。WebUI和ComfyUI是两种较为流行的操作界面选项</p>WebUI :<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">优点</strong>:界面友好,插件丰富,新手小白基本<span style="color: black;">亦</span>能秒上手</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">缺点</strong>:吃显存,对配置<span style="color: black;">需求</span>较高,出图较慢</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">2. <strong style="color: blue;">ComfyUI </strong>:<strong style="color: blue;">优点</strong>:性能好,速度快,支持工作流的导入导出分享,对小显存友好(GPU<span style="color: black;">少于</span>3G以下依然<span style="color: black;">能够</span>工作),基于工作流,对出图<span style="color: black;">规律</span>理解更清晰 </p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">缺点</strong>:对新手用户不太友好,有<span style="color: black;">必定</span>学习成本</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="//q8.itc.cn/images01/20240604/0f8c2753a2164d0ca474d9e8dd68e8c8.jpeg" style="width: 50%; margin-bottom: 20px;"></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">二者各有优缺点,<span style="color: black;">按照</span><span style="color: black;">自己</span><span style="color: black;">状况</span><span style="color: black;">选取</span><span style="color: black;">就可</span>。</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="//q6.itc.cn/images01/20240604/19cb2dd7aa4a4787bea86d9684549f6c.jpeg" style="width: 50%; margin-bottom: 20px;"></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">我为什么<span style="color: black;">选取</span>Comfyui?</strong>Comfyui的热度<span style="color: black;">连续</span><span style="color: black;">提升</span>,在谷歌搜索上,有越来越多的人<span style="color: black;">起始</span>关注Comfyui。<span style="color: black;">为何</span>会有越来越多的人关注?<span style="color: black;">由于</span><span style="color: black;">大众</span><span style="color: black;">运用</span>webui久了之后,<span style="color: black;">发掘</span>很难对生成过程有完全的<span style="color: black;">掌控</span>,<span style="color: black;">或</span>一张<span style="color: black;">照片</span>的生产不是点击几下就能完成的,它有<span style="color: black;">非常多</span>到工序:主体、配景、背景、风格、文字、特效等等<span style="color: black;">各样</span>各样的细节问题<span style="color: black;">必须</span>去<span style="color: black;">掌控</span>,<span style="color: black;">倘若</span><span style="color: black;">所有</span>采用webui,操作起来非常麻烦,<span style="color: black;">所有</span><span style="color: black;">必须</span>反复手工(尤其是要守在电脑前一个个去操作),而<strong style="color: blue;">采用Comfyui,搭建完工作流程后,只<span style="color: black;">必须</span>点击运行,<span style="color: black;">就可</span>全流程自动完成。</strong> </p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="//q6.itc.cn/images01/20240604/e5d71b075727479a92980dbefd5e0d06.jpeg" style="width: 50%; margin-bottom: 20px;"></p>
<h1 style="color: black; text-align: left; margin-bottom: 10px;"><span style="color: black;">2、</span>ComfyUI 能干啥?</h1>1. <span style="color: black;">基本</span>文生图2. <span style="color: black;">基本</span>图生图3. 真人转动漫/动漫转真人4. 线稿上色5. 老旧照片修复6. <span style="color: black;">隐匿</span>艺术字7. 改变<span style="color: black;">名人</span>姿态8. 四维彩超宝宝长相预测9. 红包封面10. 真人电子AI写真定制11. 赛博朋克风格转换12. 专属表情包13. 手机壁纸14. <span style="color: black;">更加多</span>:<span style="color: black;">这儿</span>不一一举例了,类似的玩法在网上<span style="color: black;">能够</span>看到<span style="color: black;">非常多</span>,ComfyUI只是一个<span style="color: black;">工具</span>,<span style="color: black;">详细</span><span style="color: black;">怎样</span>应用,就要依靠<span style="color: black;">自己</span>的想象力了<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="//q2.itc.cn/images01/20240604/af6cf51ba6c54674b9a9af0e2e581fe8.jpeg" style="width: 50%; margin-bottom: 20px;"></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="//q0.itc.cn/images01/20240604/e933ddb2225f4f86b2723a4d094048cd.jpeg" style="width: 50%; margin-bottom: 20px;"></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">ComfyUI 安装方式</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">1.在b 站或后期圈gzh (附带工作流)下载整合包完成后<span style="color: black;">而后</span>解压。</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="//q6.itc.cn/images01/20240604/0f87d6c7578d468fb2f68a3ac5156a8a.jpeg" style="width: 50%; margin-bottom: 20px;"></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;"><span style="color: black;">运用</span>的磁盘最好有20g以上的空间,<span style="color: black;">由于</span>再下载<span style="color: black;">有些</span>模型还是比较占用磁盘空间的。</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">“A绘世<span style="color: black;">起步</span>器.exe”文件即为一键<span style="color: black;">起步</span>文件,<span style="color: black;">能够</span>先不<span style="color: black;">起步</span>,先下载好模型。</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">下载模型,<span style="color: black;">或</span>和webui共用模型 模型文件<span style="color: black;">一般</span>以.safetensors结尾。</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;"><span style="color: black;">首要</span>将comfy ui根目录下的“extra_model_paths.yaml.example”文件重命名为“extra_model_paths.yaml”。</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="//q5.itc.cn/images01/20240604/6ec3fa062b434ad581a0944f555f7d36.jpeg" style="width: 50%; margin-bottom: 20px;"></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;"><span style="color: black;">而后</span>将base_path:后面改为你的webui目录:</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="//q3.itc.cn/images01/20240604/22b8714b5e144d14a657e9869e71fbfb.jpeg" style="width: 50%; margin-bottom: 20px;"></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">点击“A绘世<span style="color: black;">起步</span>器.exe”。<span style="color: black;">起步</span>后软件会自动更新<span style="color: black;">有些</span>文件,<span style="color: black;">通常</span>不会很久。</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="//q6.itc.cn/images01/20240604/f7523adbd6494ae0b0edb276f3f91a3a.jpeg" style="width: 50%; margin-bottom: 20px;"></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;"><span style="color: black;">而后</span>点击右下角的“一键<span style="color: black;">起步</span>”,<span style="color: black;">第1</span>次<span style="color: black;">起步</span>,可能比较慢,等一会<span style="color: black;">就可</span>,<span style="color: black;">起步</span>完成会自动打开浏览器。<span style="color: black;">倘若</span><span style="color: black;">运用</span>过程中遇到问题,<span style="color: black;">能够</span>查看左下角“<span style="color: black;">掌控</span>台”页面里的日志。</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="//q0.itc.cn/images01/20240604/7255adf92a884df9a274d0f6279b2356.jpeg" style="width: 50%; margin-bottom: 20px;"></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">设置中文 <span style="color: black;">第1</span>次打开<span style="color: black;">倘若</span>是英文,<span style="color: black;">能够</span>设置一下语言。先点击小齿轮,进入设置页面:</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="//q6.itc.cn/images01/20240604/a7a8e47536ce4eaeabc40b675295c171.jpeg" style="width: 50%; margin-bottom: 20px;"></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;"><span style="color: black;">而后</span>滑到最下面,找到:AGLTranslation-langualge选项,改为中文<span style="color: black;">就可</span>。</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="//q4.itc.cn/images01/20240604/11a52335b1134862b6ae85ad6557c16f.jpeg" style="width: 50%; margin-bottom: 20px;"><img src="//q6.itc.cn/images01/20240604/f9de21f0ee0c406b917ef0935c5db5d9.jpeg" style="width: 50%; margin-bottom: 20px;"></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">ComfyUI 工作流<span style="color: black;">基本</span>教程</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">No.1</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">文生图工作流</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">原理简介</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">1、<span style="color: black;">首要</span>将<strong style="color: blue;">prompt文本</strong>转换为<strong style="color: blue;">词特征向量</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">2、<strong style="color: blue;">词特征向量</strong>和<strong style="color: blue;">随机图向量</strong><span style="color: black;">一块</span>encode降维输入<strong style="color: blue;">潜空间</strong>,进行屡次降噪</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">输出图 = 输入图 -【(<span style="color: black;">按照</span>prompt预测的噪声+<span style="color: black;">按照</span>随机图预测的噪声-<span style="color: black;">按照</span>随机图预测的噪声)* 权重系数 +<span style="color: black;">按照</span>随机图预测的噪声】* 降噪次数</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">3、输出数据decode</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">在软件中,<span style="color: black;">已然</span>内置了许多工作流模版,下图对文生图<span style="color: black;">做为</span><span style="color: black;">基本</span>工作流的各模块进行一个简单介绍。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">下图中工作流对应文生图原理,<span style="color: black;">按照</span>Base Model生成词特征向量和随机图进行解码、屡次降噪、并编码生成<span style="color: black;">照片</span>。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="//q4.itc.cn/images01/20240604/1b70f3bc40fe4f388cdb0ffcc1218e49.jpeg" style="width: 50%; margin-bottom: 20px;"></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">ComfyUI 文生图基本工作流</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">提示词撰写</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">写提示词<span style="color: black;">必须</span><span style="color: black;">重视</span>,长度约60字,SD的提示词不像自然语言生图的<span style="color: black;">工具</span>直接组织为句子输入,输入形式为<span style="color: black;">重要</span>词<span style="color: black;">摆列</span>,越<span style="color: black;">重要</span>的越靠前,或<span style="color: black;">运用</span>权重设置如(keyword:1.4)<span style="color: black;">加强</span>权重;(keyword:0.7)降低权重</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">Prompt内容<span style="color: black;">通常</span><span style="color: black;">包括</span></strong></p>主体环境(地点,灯光,天气)作画形式(油画、水粉、素描、相机……)风格(年代、<span style="color: black;">名人</span>、艺术类型、国家……)<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">其他<span style="color: black;">重视</span>事项</strong></p>构图<span style="color: black;">意见</span><span style="color: black;">运用</span>图生图在prompt中<span style="color: black;">运用</span>别人训练好的embedding模型<span style="color: black;">能够</span>生成特定的视觉特征(某特定<span style="color: black;">名人</span>、某物种……) embedding:filename<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">尺寸设置</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">宽和高<span style="color: black;">必要</span>是8的倍数,Batch_size决定了一次生成多少张图像(可能影响速度,<span style="color: black;">意见</span>一次一张)</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><img src="//q0.itc.cn/images01/20240604/f11b4f3b05174e6f8770b4effc770025.jpeg" style="width: 50%; margin-bottom: 20px;"></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">尺寸设置参数</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">采样器参数设置</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">(参数设置<span style="color: black;">能够</span>参考模型作者<span style="color: black;">供给</span>的最佳<span style="color: black;">方法</span>)</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">Seed:</strong>随机种子数,这个数字<span style="color: black;">掌控</span>每次生成的<span style="color: black;">照片</span><span style="color: black;">类似</span>程度,<span style="color: black;">照片</span><span style="color: black;">必须</span>保持一致的时候,该数字不变</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">Control-after-generated:</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">生成完之后种子数的变化<span style="color: black;">状况</span>设置</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">Step:</strong>降噪步数</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">CFG</strong>:Prompt权重系数(<span style="color: black;">通常</span>6-8)</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">Sampler-name:</strong>采样器名<span style="color: black;">叫作</span></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">Scheduler:</strong>调度器名<span style="color: black;">叫作</span>(降噪)</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">Denoise:</strong>初始噪声(文生图<span style="color: black;">通常</span>为1)</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">Lora配置</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">添加lora模块,更改clip、model连线,更换作者<span style="color: black;">举荐</span>的Vae模块等其他参数</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">添加模块</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">起步</span></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">此时设置完所有的模型和参数,<span style="color: black;">根据</span>教程的参数设置运行后依然会<span style="color: black;">发掘</span>出图很糊,<span style="color: black;">据述</span>M系列Mac容易遇到该问题,<span style="color: black;">按照</span>教程,在终端pip安装 torchvision==0.16.2,成功<span style="color: black;">处理</span>该问题。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">打开终端进行pip安装,输入 torchvision==0.16.2</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">处理</span><span style="color: black;">照片</span>输出模糊的问题</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">No.2</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">图生图工作流</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;"><span style="color: black;">办法</span>一:重绘 - 基于输入图进行重新绘制</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">输入图代替随机图,和文字<span style="color: black;">一块</span><span style="color: black;">做为</span>潜空间的输入。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;"><span style="color: black;">办法</span>二:参考 - 基于输入图参考进行内容<span style="color: black;">弥补</span>(unCLIP)、风格迁移(Style)拓展</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">先对图文进行解码,<span style="color: black;">而后</span>和随机图<span style="color: black;">一块</span>输入潜空间</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">unCLIP model workflow</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">针对</span>unCLIP重绘方式,<span style="color: black;">能够</span>直接<span style="color: black;">运用</span>模版搭好的工作流,其中<span style="color: black;">增多</span>了上图所示三个模块,并将初始CheckpointLoader、对应的unclip模型进行替换(模型:sd2.1-unclip, 对应随机图尺寸 768 * 768)</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">当<span style="color: black;">运用</span>两张图进行元素融合,<span style="color: black;">能够</span>看到结果大致实现将鸭子和水波元素进行了混合,但与预期效果差距<span style="color: black;">很强</span>,但融合的元素难以<span style="color: black;">掌控</span>,若要<span style="color: black;">加强</span>符合预期的质量仍然<span style="color: black;">必须</span>进行<span style="color: black;">更加多</span>的参数<span style="color: black;">调节</span>。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">初始输入<span style="color: black;">照片</span>高度<span style="color: black;">调节</span>为768 跑unCLIP工作流</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">Style model workflow</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">风格迁移工作流中则<span style="color: black;">必须</span>将上传的<span style="color: black;">照片</span>编码为向量、理解风格、理解提示词,<span style="color: black;">经过</span>Style节点处理后导入采样器,<span style="color: black;">通常</span>为名画、<span style="color: black;">名人</span>雕塑类。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">跑Style Model Workflow</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">此处<span style="color: black;">运用</span>了新海城画风的<span style="color: black;">照片</span>输入,整体而言,风格更加依赖模型,<span style="color: black;">一样</span>参数下,模型AnythingElse为二次元专用生<span style="color: black;">成为了</span>左边的效果,DreamShaper为写实风生<span style="color: black;">成为了</span>右边的效果,输出图风格<span style="color: black;">更加多</span>与模型风格、prompt内容更<span style="color: black;">关联</span>,与上传的风格迁移的参考<span style="color: black;">照片</span>关系不大。</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">AnythingElse(左),DreamShaper(右)</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><strong style="color: blue;">ComfyUI 工作流分享(<span style="color: black;">照片</span>水印)</strong></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">6 月工作流更新</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;">AI模特、AI换脸、AI抠图、AI制作PPT、AI音频、AI视频、AI物体消除</p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><a style="color: black;"><span style="color: black;">返回<span style="color: black;">外链论坛:http://www.fok120.com/</span>,查看<span style="color: black;">更加多</span></span></a></p>
<p style="font-size: 16px; color: black; line-height: 40px; text-align: left; margin-bottom: 15px;"><span style="color: black;">责任编辑:网友投稿</span></p>
你的话语真是温暖如春,让我心生感激。 太棒了、厉害、为你打call、点赞、非常精彩等。 我深受你的启发,你的话语是我前进的动力。
页:
[1]