testforpptxrag 7.17MB

shuaikang9864需要积分:1(1积分=1元)

资源文件列表:

rag调研.pptx.zip 大约有2个文件

rag调研.pptx 7.52MB
__MACOSX/._rag调研.pptx 375B

资源介绍:

testforpptxrag

<link href="/image.php?url=https://csdnimg.cn/release/download_crawler_static/css/base.min.css" rel="stylesheet"/><link href="/image.php?url=https://csdnimg.cn/release/download_crawler_static/css/fancy.min.css" rel="stylesheet"/><link href="/image.php?url=https://csdnimg.cn/release/download_crawler_static/89506981/raw.css" rel="stylesheet"/><div id="sidebar" style="display: none"><div id="outline"></div></div><div class="pf w0 h0" data-page-no="1" id="pf1"><div class="pc pc1 w0 h0"><img alt="" class="bi x0 y0 w1 h1" src="/image.php?url=https://csdnimg.cn/release/download_crawler_static/89506981/bg1.jpg"/><div class="t m0 x1 h2 y1 ff1 fs0 fc0 sc0 ls0 ws0">项目</div><div class="t m0 x2 h2 y2 ff2 fs0 fc0 sc0 ls0 ws0">pdf Parser<span class="_ _0"> </span><span class="ff1">其他</span>Parser</div><div class="t m0 x3 h2 y1 ff1 fs0 fc0 sc0 ls0 ws0">存储结构</div><div class="t m0 x4 h2 y2 ff2 fs0 fc0 sc0 ls0 ws0">retrieval<span class="_ _1"> </span>LLM<span class="ff1">调用</span></div><div class="t m0 x5 h2 y1 ff1 fs0 fc0 sc0 ls0 ws0">其他方面</div><div class="t m0 x1 h3 y3 ff3 fs1 fc1 sc0 ls0 ws0">RAGFlow</div><div class="t m0 x1 h3 y4 ff3 fs1 fc1 sc0 ls0 ws0">(python-</div><div class="t m0 x1 h3 y5 ff3 fs1 fc1 sc0 ls0 ws0">TypeScript</div><div class="t m0 x1 h3 y6 ff3 fs1 fc1 sc0 ls0 ws0">)</div><div class="t m0 x2 h3 y3 ff3 fs1 fc1 sc0 ls0 ws0">pyPDF2+ocr</div><div class="t m0 x2 h4 y4 ff4 fs1 fc1 sc0 ls0 ws0">优<span class="ff5">化了中文解析</span></div><div class="t m0 x2 h4 y5 ff5 fs1 fc1 sc0 ls0 ws0">正<span class="ff4">则</span>化</div><div class="t m0 x6 h5 y3 ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">markdown</span></div><div class="t m0 x6 h5 y4 ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">pptx,docx</span></div><div class="t m0 x6 h5 y5 ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">html_text, </span></div><div class="t m0 x7 h3 y6 ff3 fs1 fc1 sc0 ls0 ws0">readability</div><div class="t m0 x6 h5 y7 ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">chardet</span></div><div class="t m0 x6 h5 y8 ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">openpyxl</span></div><div class="t m0 x6 h5 y9 ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">deepdoc:ocr+ts</span></div><div class="t m0 x7 h3 ya ff3 fs1 fc1 sc0 ls0 ws0">r</div><div class="t m0 x3 h3 y3 ff3 fs1 fc1 sc0 ls0 ws0">minio->FS</div><div class="t m0 x3 h6 y4 ff3 fs1 fc1 sc0 ls0 ws0">mysql-><span class="ff5">基本信息</span></div><div class="t m0 x3 h4 y5 ff3 fs1 fc1 sc0 ls0 ws0">redis-><span class="ff5">向量存<span class="ff4">储</span>，</span>chunks</div><div class="t m0 x3 h5 y6 ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">Documents (PDF, DOCS)</span></div><div class="t m0 x3 h5 y7 ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">Tables (XLSX)</span></div><div class="t m0 x3 h5 y8 ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">Pictures (JPEG, JPG, PNG, </span></div><div class="t m0 x8 h3 y9 ff3 fs1 fc1 sc0 ls0 ws0">TIF, GIF)</div><div class="t m0 x3 h4 ya ff5 fs1 fc1 sc0 ls0 ws0">普通：<span class="ff3">chunks</span>，<span class="ff3">qa<span class="ff4">对</span></span>，表格，</div><div class="t m0 x3 h4 yb ff3 fs1 fc1 sc0 ls0 ws0">laws<span class="ff5">，</span>ppt<span class="ff5">，一整个文件直接<span class="ff4">储</span></span></div><div class="t m0 x3 h6 yc ff5 fs1 fc1 sc0 ls0 ws0">存方式</div><div class="t m0 x4 h6 y3 ff5 fs1 fc1 sc0 ls0 ws0">本地部署支持：</div><div class="t m0 x4 h6 y4 ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">bge<span class="ff5">全系列</span></span></div><div class="t m0 x4 h5 y5 ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">jina-embedding-v2</span></div><div class="t m0 x4 h5 y6 ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">nomic-embed-text</span></div><div class="t m0 x4 h5 y7 ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">all-MiniLM-L6-v2</span></div><div class="t m0 x4 h5 y8 ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">bce-embedding-base_v1</span></div><div class="t m0 x9 h6 y3 ff3 fs1 fc1 sc0 ls0 ws0">ollama<span class="ff5">和</span></div><div class="t m0 x9 h6 y4 ff3 fs1 fc1 sc0 ls0 ws0">inference<span class="ff5">部署</span></div><div class="t m0 x5 h4 y3 ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="ff4 fs1">细<span class="ff5">粒度解析和文档溯源</span></span></div><div class="t m0 xa h4 y4 ff5 fs1 fc1 sc0 ls0 ws0">展示做的比<span class="ff4">较</span>好。</div><div class="t m0 x5 h4 y5 ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="ff5 fs1">目前支持英<span class="ff4">语</span>、<span class="ff4">简</span>体中</span></div><div class="t m0 xa h6 y6 ff5 fs1 fc1 sc0 ls0 ws0">文和繁体中文。</div><div class="t m0 x5 h6 y7 ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="ff5 fs1">内置文档解析模型</span></div><div class="t m0 x1 h3 yd ff3 fs1 fc1 sc0 ls0 ws0">LangFlow</div><div class="t m0 x1 h3 ye ff3 fs1 fc1 sc0 ls0 ws0">(JS-</div><div class="t m0 x1 h3 yf ff3 fs1 fc1 sc0 ls0 ws0">TypeScript</div><div class="t m0 x1 h3 y10 ff3 fs1 fc1 sc0 ls0 ws0">)</div><div class="t m0 x2 h4 yd ff3 fs1 fc1 sc0 ls0 ws0">react-pdf<span class="_ _3"> </span><span class="ff5">正在开<span class="ff4">发图</span>像、音</span></div><div class="t m0 x6 h4 ye ff4 fs1 fc1 sc0 ls0 ws0">频<span class="ff5">、</span>视频<span class="ff5">等解析</span>储</div><div class="t m0 x6 h6 yf ff5 fs1 fc1 sc0 ls0 ws0">存</div><div class="t m0 x6 h5 y10 ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">code_parser</span></div><div class="t m0 x6 h5 y11 ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">image: </span></div><div class="t m0 x7 h3 y12 ff3 fs1 fc1 sc0 ls0 ws0">OpenSeadrago</div><div class="t m0 x7 h3 y13 ff3 fs1 fc1 sc0 ls0 ws0">n</div><div class="t m0 x3 h6 yd ff3 fs1 fc1 sc0 ls0 ws0">Astra DB, postgre->chunks<span class="ff5">，向</span></div><div class="t m0 x3 h6 ye ff5 fs1 fc1 sc0 ls0 ws0">量</div><div class="t m0 x3 h6 yf ff5 fs1 fc1 sc0 ls0 ws0">支持<span class="ff3">chromadb</span></div><div class="t m0 x4 h5 yd ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">Amazon Bedrock </span></div><div class="t m0 xb h3 ye ff3 fs1 fc1 sc0 ls0 ws0">Embeddings</div><div class="t m0 x4 h5 yf ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">Astra vectoriz</span></div><div class="t m0 x4 h5 y10 ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">Cohere Embeddings</span></div><div class="t m0 x4 h5 y11 ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">OpenAI Embeddings</span></div><div class="t m0 x4 h6 y12 ff5 fs1 fc1 sc0 ls0 ws0">本地</div><div class="t m0 x4 h5 y13 ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">Ollama Embeddings</span></div><div class="t m0 x4 h5 y14 ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">VertexAI Embeddings</span></div><div class="t m0 x4 h5 y15 ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">Hugging Face Embeddings</span></div><div class="t m0 x9 h6 yd ff5 fs1 fc1 sc0 ls0 ws0">本地：</div><div class="t m0 x9 h5 ye ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">Qianfan</span></div><div class="t m0 x9 h5 yf ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">Vertex AI</span></div><div class="t m0 x9 h5 y10 ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">Ollama</span></div><div class="t m0 x5 h4 yd ff5 fs1 fc1 sc0 ls0 ws0">用<span class="ff4">户</span>可以<span class="ff4">编辑</span>或<span class="ff4">删</span>除之前</div><div class="t m0 x5 h4 ye ff5 fs1 fc1 sc0 ls0 ws0">的<span class="ff4">问</span>答内容<span class="ff4">给</span>后面的<span class="ff4">对话</span></div><div class="t m0 x5 h6 yf ff5 fs1 fc1 sc0 ls0 ws0">做可控<span class="ff3">prompt</span></div><div class="t m0 x5 h6 y10 ff5 fs1 fc1 sc0 ls0 ws0">更方便构建<span class="ff3">agent</span>工作流</div><div class="t m0 x1 h3 y16 ff3 fs1 fc1 sc0 ls0 ws0">FastGPT</div><div class="t m0 x1 h3 y17 ff3 fs1 fc1 sc0 ls0 ws0">(TypeScrip</div><div class="t m0 x1 h3 y18 ff3 fs1 fc1 sc0 ls0 ws0">t-JS)</div><div class="t m0 x2 h3 y16 ff3 fs1 fc1 sc0 ls0 ws0">pdfjs</div><div class="t m0 x2 h4 y17 ff5 fs1 fc1 sc0 ls0 ws0">只能<span class="ff4">识别</span>文字</div><div class="t m0 x6 h6 y16 ff5 fs1 fc1 sc0 ls0 ws0">可以解析<span class="ff3">word</span>，</div><div class="t m0 x6 h6 y17 ff3 fs1 fc1 sc0 ls0 ws0">excel<span class="ff5">，可以构建</span></div><div class="t m0 x6 h6 y18 ff5 fs1 fc1 sc0 ls0 ws0">表</div><div class="t m0 x3 h4 y16 ff3 fs1 fc1 sc0 ls0 ws0">MongoDB->FS<span class="ff4">储<span class="ff5">存</span></span></div><div class="t m0 x3 h4 y17 ff3 fs1 fc1 sc0 ls0 ws0">Postgre->chunk<span class="ff5">，向量<span class="ff4">储</span>存</span></div><div class="t m0 x3 h3 y18 ff3 fs1 fc1 sc0 ls0 ws0">Mysql->openapi</div><div class="t m0 x3 h4 y19 ff5 fs1 fc1 sc0 ls0 ws0">在上<span class="ff4">传</span>知<span class="ff4">识库处</span>理<span class="ff4">时</span>，就需要<span class="ff4">进</span></div><div class="t m0 x3 h4 y1a ff5 fs1 fc1 sc0 ls0 ws0">行分<span class="ff4">类储</span>存管理：支持直接</div><div class="t m0 x3 h4 y1b ff3 fs1 fc1 sc0 ls0 ws0">chunk<span class="ff5">，</span>qa<span class="ff4">对处<span class="ff5">理，外部</span></span>url<span class="_ _4"></span><span class="ff5">知<span class="ff4">识</span></span></div><div class="t m0 x3 h4 y1c ff4 fs1 fc1 sc0 ls0 ws0">库<span class="ff5">，高</span>级<span class="ff5">摘要索引</span>转<span class="ff5">化功能</span></div><div class="t m0 x4 h6 y16 ff3 fs1 fc1 sc0 ls0 ws0">embedding model<span class="_ _4"></span><span class="ff5">支持：</span></div><div class="t m0 x4 h4 y17 ff5 fs1 fc1 sc0 ls0 ws0">内置的<span class="ff3">embedding api<span class="_ _4"></span><span class="ff5">接口<span class="ff4">较</span></span></span></div><div class="t m0 x4 h6 y18 ff5 fs1 fc1 sc0 ls0 ws0">少</div><div class="t m0 x4 h6 y19 ff5 fs1 fc1 sc0 ls0 ws0">本地部署支持<span class="ff3">m3e</span>，<span class="ff3">bge</span></div><div class="t m0 x9 h4 y16 ff5 fs1 fc1 sc0 ls0 ws0">通<span class="ff4">过<span class="ff3">openapi<span class="_ _4"></span><span class="ff5">控</span></span></span></div><div class="t m0 x9 h6 y17 ff5 fs1 fc1 sc0 ls0 ws0">制</div><div class="t m0 x9 h6 y18 ff5 fs1 fc1 sc0 ls0 ws0">可以兼容<span class="ff3">ollama</span></div><div class="t m0 x9 h4 y19 ff5 fs1 fc1 sc0 ls0 ws0">和<span class="ff3">docker</span>端口<span class="ff4">访</span></div><div class="t m0 x9 h4 y1a ff4 fs1 fc1 sc0 ls0 ws0">问<span class="ff5">模式</span></div><div class="t m0 x5 h4 y16 ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="ff5 fs1">技<span class="ff4">术栈</span>：<span class="ff3">NextJs + </span></span></div><div class="t m0 xa h3 y17 ff3 fs1 fc1 sc0 ls0 ws0">TypeScript + ChakraUI</div><div class="t m0 x5 h4 y18 ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="ff4 fs1">强调<span class="ff5">自己有独特的<span class="ff3">qa</span></span>结</span></div><div class="t m0 xa h4 y19 ff5 fs1 fc1 sc0 ls0 ws0">构，适合客服<span class="ff4">问</span>答<span class="ff4">场</span>景</div><div class="t m0 x5 h4 y1a ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="ff5 fs1">点<span class="ff4">击</span>提示需要付<span class="ff4">费</span>的高</span></div><div class="t m0 xa h4 y1b ff4 fs1 fc1 sc0 ls0 ws0">级<span class="ff5">功能按</span>钮<span class="ff5">太多</span></div><div class="t m0 x1 h3 y1d ff3 fs1 fc1 sc0 ls0 ws0">DocsGPT(</div><div class="t m0 x1 h3 y1e ff3 fs1 fc1 sc0 ls0 ws0">python-</div><div class="t m0 x1 h3 y1f ff3 fs1 fc1 sc0 ls0 ws0">TypeScript</div><div class="t m0 x1 h3 y20 ff3 fs1 fc1 sc0 ls0 ws0">)</div><div class="t m0 x2 h5 y1d ff3 fs1 fc1 sc0 ls0 ws0">pyPDF2<span class="_ _5"> </span><span class="fs2">•<span class="_ _2"> </span></span>epub_parser</div><div class="t m0 x6 h5 y1e ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">rst_parser</span></div><div class="t m0 x6 h5 y1f ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">markdown_par</span></div><div class="t m0 x7 h3 y20 ff3 fs1 fc1 sc0 ls0 ws0">ser</div><div class="t m0 x6 h5 y21 ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">openapi3_pars</span></div><div class="t m0 x7 h3 y22 ff3 fs1 fc1 sc0 ls0 ws0">er</div><div class="t m0 x6 h5 y23 ff3 fs2 fc1 sc0 ls0 ws0">•<span class="_ _2"> </span><span class="fs1">csv&pandas_p</span></div><div class="t m0 x7 h3 y24 ff3 fs1 fc1 sc0 ls0 ws0">arser</div><div class="t m0 x3 h4 y1d ff3 fs1 fc1 sc0 ls0 ws0">MongoDB->FS<span class="ff4">储<span class="ff5">存</span></span></div><div class="t m0 x3 h4 y1e ff3 fs1 fc1 sc0 ls0 ws0">redis-><span class="ff5">向量存<span class="ff4">储</span>，</span>chunks</div><div class="t m0 x4 h4 y1d ff5 fs1 fc1 sc0 ls0 ws0">本地部署<span class="ff3">embedding<span class="_ _4"></span><span class="ff5">模型<span class="ff4">暂时</span></span></span></div><div class="t m0 x4 h6 y1e ff5 fs1 fc1 sc0 ls0 ws0">位置</div><div class="t m0 x9 h4 y1d ff5 fs1 fc1 sc0 ls0 ws0">定制<span class="ff4">优</span>化的<span class="ff3">llm</span>：</div><div class="t m0 x9 h3 y1e ff3 fs1 fc1 sc0 ls0 ws0">Docsgpt-7b-</div><div class="t m0 x9 h3 y1f ff3 fs1 fc1 sc0 ls0 ws0">mistral</div><div class="t m0 x9 h3 y20 ff3 fs1 fc1 sc0 ls0 ws0">Docsgpt-14b</div><div class="t m0 x9 h3 y21 ff3 fs1 fc1 sc0 ls0 ws0">Docsgpt-40b-</div><div class="t m0 x9 h3 y22 ff3 fs1 fc1 sc0 ls0 ws0">falcon</div><div class="t m0 x5 h4 y1d ff4 fs1 fc1 sc0 ls0 ws0">对<span class="ff5">代</span>码<span class="ff5">的，面向工程开</span>发</div><div class="t m0 x5 h4 y1e ff5 fs1 fc1 sc0 ls0 ws0">文档的解析能力<span class="ff4">强</span></div><div class="t m0 x5 h6 y1f ff5 fs1 fc1 sc0 ls0 ws0">有<span class="ff3">py2doc</span>，<span class="ff3">js2doc<span class="_ _4"></span><span class="ff5">，</span></span></div><div class="t m0 x5 h3 y20 ff3 fs1 fc1 sc0 ls0 ws0">java2doc</div></div><div class="pi" data-data='{"ctm":[1.000000,0.000000,0.000000,1.000000,0.000000,0.000000]}'></div></div><div id="pf2" class="pf w0 h0" data-page-no="2"><div class="pc pc2 w0 h0"><img class="bi x0 y0 w1 h1" alt="" src="/image.php?url=https://csdnimg.cn/release/download_crawler_static/89506981/bg2.jpg"><div class="t m0 xc h7 y25 ff1 fs3 fc1 sc0 ls0 ws0">其他项目纵览</div><div class="t m0 xd h8 y26 ff2 fs0 fc0 sc0 ls0 ws0">Project<span class="_ _6"> </span>Advantages<span class="_ _7"> </span>Disadvantages<span class="_ _8"> </span>Deployment<span class="_ _9"> </span>Overhead</div><div class="t m0 xd h9 y27 ff3 fs4 fc1 sc0 ls0 ws0">MaxKB<span class="_ _a"> </span><span class="ff5">功能清晰<span class="ff4">简洁</span></span></div><div class="t m0 xe ha y28 ff5 fs4 fc1 sc0 ls0 ws0">支持中文</div><div class="t m0 xf h9 y27 ff5 fs4 fc1 sc0 ls0 ws0">不<span class="ff4">显</span>示引用知<span class="ff4">识库</span>原文</div><div class="t m0 xf h9 y28 ff5 fs4 fc1 sc0 ls0 ws0">没有提示是否命中知<span class="ff4">识库</span></div><div class="t m0 x10 h9 y27 ff4 fs4 fc1 sc0 ls0 ws0">简单<span class="_ _b"> </span><span class="ff5">低</span></div><div class="t m0 xd h9 y29 ff3 fs4 fc1 sc0 ls0 ws0">AnythingLLM<span class="_ _c"> </span><span class="ff5">可<span class="ff4">选</span></span>chat/query<span class="ff5">模式</span></div><div class="t m0 xe h9 y2a ff3 fs4 fc1 sc0 ls0 ws0">query<span class="ff5">模式每次<span class="ff4">查询</span></span>KB</div><div class="t m0 xf h9 y29 ff5 fs4 fc1 sc0 ls0 ws0">不支持中文<span class="_ _d"> </span><span class="ff4">简单<span class="_ _b"> </span></span>低</div><div class="t m0 xd h9 y2b ff3 fs4 fc1 sc0 ls0 ws0">Dify.ai<span class="_ _e"> </span><span class="ff5">支持中文，界面清晰<span class="_ _f"> </span>不<span class="ff4">显</span>示引用原文</span></div><div class="t m0 xf ha y2c ff5 fs4 fc1 sc0 ls0 ws0">不提示是否命中<span class="ff3">kb</span></div><div class="t m0 x10 ha y2b ff5 fs4 fc1 sc0 ls0 ws0">中等<span class="_ _b"> </span>中</div><div class="t m0 xd h9 y2d ff3 fs4 fc1 sc0 ls0 ws0">FastGPT<span class="_ _10"> </span><span class="ff5">支持</span>1000<span class="ff5">个文件，<span class="ff4">单</span>文</span></div><div class="t m0 xe ha y2e ff5 fs4 fc1 sc0 ls0 ws0">件最大<span class="ff3">500mb</span></div><div class="t m0 xe ha y2f ff5 fs4 fc1 sc0 ls0 ws0">命中率高，提示原文</div><div class="t m0 xf h9 y2d ff5 fs4 fc1 sc0 ls0 ws0">本地<span class="ff3">KB</span>配置<span class="ff4">难</span>度高<span class="_ _11"> </span><span class="ff4">较难<span class="_ _b"> </span></span>中</div><div class="t m0 xd h9 y30 ff3 fs4 fc1 sc0 ls0 ws0">RAGFlow<span class="_ _12"> </span><span class="ff5">自<span class="ff4">带</span>大模型</span></div><div class="t m0 xe h9 y31 ff5 fs4 fc1 sc0 ls0 ws0">知<span class="ff4">识库</span>命中率高、<span class="ff4">显</span>示原</div><div class="t m0 xe ha y32 ff5 fs4 fc1 sc0 ls0 ws0">文</div><div class="t m0 xf h9 y30 ff5 fs4 fc1 sc0 ls0 ws0">体<span class="ff4">积</span>太大<span class="_ _13"> </span><span class="ff4">较难<span class="_ _b"> </span></span>高</div></div><div class="pi" data-data='{"ctm":[1.000000,0.000000,0.000000,1.000000,0.000000,0.000000]}'></div></div><div id="pf3" class="pf w0 h0" data-page-no="3"><div class="pc pc3 w0 h0"><img class="bi x0 y0 w1 h1" alt="" src="/image.php?url=https://csdnimg.cn/release/download_crawler_static/89506981/bg3.jpg"><div class="t m0 xc hb y33 ff6 fs3 fc1 sc0 ls0 ws0">RAGFlow</div></div><div class="pi" data-data='{"ctm":[1.000000,0.000000,0.000000,1.000000,0.000000,0.000000]}'></div></div>

100+评论昵称:

captcha

类型	标题	大小	时间
	web_QQ浏览器压缩包.zip	163.62KB	10月前

	fgsjkhghILafsaws	13.03MB	10月前
	DWinterTabDemo.zip	396.48KB	12年前
	win11环境下VMware Workstation pro运行虚拟机蓝屏修复	369B	3年前
	2001292440网络234韩艺zq.zip	9.15MB	10月前

	XCP协议及文档及A2L文件	39.59MB	10月前
	linux 下生成图片的dll	908.28KB	10月前
	springboot‘s taotal test	40.75KB	10月前