https://www.leanku.com/tags/ragflow/Recent content in RAGFlow on Leanku 的博客Hugozh-cnTue, 28 Oct 2025 00:01:01 +0800https://www.leanku.com/post/ai/ragflow/Tue, 28 Oct 2025 00:01:01 +0800https://www.leanku.com/post/ai/ragflow/<h1 id="ragflow">RAGFlow</h1> <h2 id="一介绍">一、介绍</h2> <p>RAGFlow是一个完整的、开源的RAG（检索增强生成）系统。你可以把它看作一个已经搭建好的智能问答应用框架。</p> <p>RAGFlow不是一个需要你从零编码的库，而是一个可以直接运行起来的服务。它内置了我们聊过的RAG的完整流程：上传文档 -&gt; 自动解析 -&gt; 切片 -&gt; 调用Embedding模型生成向量 -&gt; 存储 -&gt; 检索 -&gt; 调用大模型生成回答。</p> <p><a href="https://ragflow.io/">RAGFlow 官网</a></p> <h2 id="二搭建一个智能问答系统">二、搭建一个智能问答系统</h2> <p>整体架构：</p> <div class="highlight"><pre tabindex="0" class="chroma"><code class="language-text" data-lang="text"><span class="line"><span class="ln">1</span><span class="cl">你的文档 (PDF/Word/PPT) </span></span><span class="line"><span class="ln">2</span><span class="cl"> ↓ </span></span><span class="line"><span class="ln">3</span><span class="cl">[RAGFlow核心] </span></span><span class="line"><span class="ln">4</span><span class="cl"> ├─ 深度文档解析 → 保留表格/图片/格式 </span></span><span class="line"><span class="ln">5</span><span class="cl"> ├─ 智能分块 → 按语义切割文档 </span></span><span class="line"><span class="ln">6</span><span class="cl"> ├─ 向量化 → 调用Embedding模型 </span></span><span class="line"><span class="ln">7</span><span class="cl"> └─ 向量存储 → 存入数据库 </span></span><span class="line"><span class="ln">8</span><span class="cl"> ↓ </span></span><span class="line"><span class="ln">9</span><span class="cl">用户提问 → [检索引擎] → [你的微调模型] → 带引用的精准回答 </span></span></code></pre></div><h3 id="21-部署ragflow">2.1 部署RAGFlow</h3> <p>这里使用 <strong>Docker</strong> 方式 环境要求：</p> <ul> <li>CPU ≥ 4 cores (x86);</li> <li>RAM ≥ 16 GB;</li> <li>Disk ≥ 50 GB;</li> <li>Docker ≥ 24.0.0 &amp; Docker Compose ≥ v2.26.1.</li> <li><a href="https://gvisor.dev/docs/user_guide/install/">gVisor</a>: Required only if you intend to use the code executor (<a href="https://github.com/infiniflow/ragflow/tree/main/sandbox">sandbox</a>) feature of RAGFlow.</li> </ul> <p><strong>部署RAGFlow</strong></p>