{"id":2546,"date":"2025-07-08T15:05:09","date_gmt":"2025-07-08T07:05:09","guid":{"rendered":"https:\/\/sanlangcode.com\/?p=2546"},"modified":"2025-07-08T15:05:09","modified_gmt":"2025-07-08T07:05:09","slug":"ragflow%ef%bc%9a%e4%b8%80%e7%ab%99%e5%bc%8f%e6%99%ba%e8%83%bdrag%e5%b9%b3%e5%8f%b0%e7%9a%84%e6%9e%b6%e6%9e%84%e6%8f%ad%e7%a7%98","status":"publish","type":"post","link":"https:\/\/sanlangcode.com\/index.php\/2025\/07\/08\/ragflow%ef%bc%9a%e4%b8%80%e7%ab%99%e5%bc%8f%e6%99%ba%e8%83%bdrag%e5%b9%b3%e5%8f%b0%e7%9a%84%e6%9e%b6%e6%9e%84%e6%8f%ad%e7%a7%98\/","title":{"rendered":"Ragflow\uff1a\u4e00\u7ad9\u5f0f\u667a\u80fdRAG\u5e73\u53f0\u7684\u67b6\u6784\u63ed\u79d8"},"content":{"rendered":"\n
\n

\u201c\u4ee3\u7801\u5982\u8bd7\uff0c\u67b6\u6784\u5982\u753b\u3002\u4eca\u5929\u6211\u4eec\u6765\u804a\u804a ragflow-main \u8fd9\u5e45\u2018\u6e05\u660e\u4e0a\u6cb3\u56fe\u2019\u3002\u201d<\/p>\n<\/blockquote>\n\n\n\n

\u4e00\u3001\u9879\u76ee\u6982\u89c8<\/h2>\n\n\n\n

ragflow-main \u662f\u4e00\u4e2a\u96c6\u6210\u4e86 RAG\uff08Retrieval-Augmented Generation\uff0c\u68c0\u7d22\u589e\u5f3a\u751f\u6210\uff09\u3001\u591a\u667a\u80fd\u4f53\u3001\u77e5\u8bc6\u5e93\u3001\u63d2\u4ef6\u3001\u6c99\u7bb1\u6267\u884c\u3001SDK\u3001Web \u524d\u7aef\u7b49\u4e00\u7cfb\u5217\u529f\u80fd\u7684 AI \u5e73\u53f0\u3002\u5b83\u4e0d\u4ec5\u80fd\u8ba9\u5927\u6a21\u578b\u201c\u6709\u8111\u6709\u8bb0\u6027\u201d\uff0c\u8fd8\u80fd\u8ba9\u4f60\u50cf\u642d\u79ef\u6728\u4e00\u6837\u6269\u5c55\u529f\u80fd\u3002<\/p>\n\n\n\n

\u9879\u76ee\u7ed3\u6784\u5927\u81f4\u5982\u4e0b\uff1a<\/p>\n\n\n\n

ragflow-main\/\n  \u251c\u2500\u2500 agent\/                # \u667a\u80fd\u4f53\u6838\u5fc3\u4e0e\u7ec4\u4ef6\n  \u251c\u2500\u2500 agentic_reasoning\/    # \u591a\u667a\u80fd\u4f53\u63a8\u7406\n  \u251c\u2500\u2500 api\/                  # \u540e\u7aefAPI\u670d\u52a1\n  \u251c\u2500\u2500 web\/                  # \u524d\u7aefReact\u5e94\u7528\n  \u251c\u2500\u2500 plugin\/               # \u63d2\u4ef6\u673a\u5236\n  \u251c\u2500\u2500 rag\/                  # RAG\u6838\u5fc3\u4e0eNLP\n  \u251c\u2500\u2500 deepdoc\/              # \u6587\u6863\u89e3\u6790\u4e0eOCR\n  \u251c\u2500\u2500 graphrag\/             # \u56fe\u8c31RAG\n  \u251c\u2500\u2500 sandbox\/              # \u6c99\u7bb1\u4ee3\u7801\u6267\u884c\n  \u251c\u2500\u2500 sdk\/                  # Python SDK\n  \u251c\u2500\u2500 conf\/\u3001docker\/\u3001docs\/ # \u914d\u7f6e\u3001\u90e8\u7f72\u3001\u6587\u6863\n  \u2514\u2500\u2500 test\/                 # \u6d4b\u8bd5<\/code><\/pre>\n\n\n\n
\u4e8c\u3001\u6280\u672f\u9009\u578b<\/h2>\n\n\n\n
    \n
  • \u540e\u7aef<\/strong>\uff1aPython 3.x\uff0cFastAPI\uff08api\/\uff09\uff0cSQLAlchemy\uff0c\u4f9d\u8d56\u4e30\u5bcc\u7684 NLP\/AI \u751f\u6001<\/li>\n\n\n\n
  • \u524d\u7aef<\/strong>\uff1aReact + TypeScript\uff08web\/\uff09\uff0cAnt Design\uff0cUmiJS<\/li>\n\n\n\n
  • RAG\/NLP<\/strong>\uff1a\u81ea\u7814 RAG pipeline\uff0c\u652f\u6301\u591a\u79cd LLM\uff0c\u5206\u8bcd\u3001NER\u3001\u77e5\u8bc6\u5e93<\/li>\n\n\n\n
  • \u63d2\u4ef6<\/strong>\uff1aPython \u52a8\u6001\u52a0\u8f7d\uff0c\u652f\u6301 LLM \u5de5\u5177\u94fe<\/li>\n\n\n\n
  • \u6c99\u7bb1<\/strong>\uff1a\u652f\u6301 Python\/Nodejs \u4ee3\u7801\u5b89\u5168\u6267\u884c<\/li>\n\n\n\n
  • SDK<\/strong>\uff1aPython \u5c01\u88c5\uff0c\u65b9\u4fbf\u4e8c\u6b21\u5f00\u53d1<\/li>\n\n\n\n
  • \u90e8\u7f72<\/strong>\uff1aDocker\u3001Helm\u3001Shell \u811a\u672c<\/li>\n\n\n\n
  • \u6587\u6863<\/strong>\uff1a\u591a\u8bed\u8a00 README\uff0c\u8be6\u7ec6\u5f00\u53d1\u8005\u6587\u6863<\/li>\n<\/ul>\n\n\n\n
    \u4e09\u3001\u6a21\u5757\u8be6\u89e3<\/h2>\n\n\n\n
    1. agent\/ \u2014\u2014 \u667a\u80fd\u4f53\u5927\u672c\u8425<\/h3>\n\n\n\n
    \u8fd9\u91cc\u662f\u201cAI\u5c0f\u5206\u961f\u201d\u7684\u603b\u90e8\u3002\u6bcf\u4e2a component\/ \u4e0b\u7684 py \u6587\u4ef6\u5c31\u662f\u4e00\u4e2a\u201c\u7279\u5de5\u201d\uff0c\u6bd4\u5982 akshare.py\uff08\u8d22\u7ecf\u6570\u636e\uff09\u3001answer.py\uff08\u95ee\u7b54\uff09\u3001search.py\uff08\u641c\u7d22\uff09\u7b49\u3002\u4f60\u53ef\u4ee5\u50cf\u62fc\u4e50\u9ad8\u4e00\u6837\u7ec4\u5408\u8fd9\u4e9b\u7ec4\u4ef6\uff0c\u7ec4\u88c5\u51fa\u81ea\u5df1\u7684\u8d85\u7ea7\u667a\u80fd\u4f53\u3002<\/p>\n\n\n\n
      \n
    • canvas.py<\/strong>\uff1a\u753b\u5e03\u5f0f\u667a\u80fd\u4f53\u7f16\u6392\uff0c\u652f\u6301 DSL\uff08domain-specific language\uff09\u5b9a\u4e49 agent \u6d41\u7a0b\u3002<\/li>\n\n\n\n
    • templates\/<\/strong>\uff1a\u9884\u8bbe\u7684 agent \u6d41\u7a0b\u6a21\u677f\uff0c\u5f00\u7bb1\u5373\u7528\u3002<\/li>\n\n\n\n
    • test\/<\/strong>\uff1aagent \u76f8\u5173\u7684\u6d4b\u8bd5\u548c DSL \u793a\u4f8b\u3002<\/li>\n<\/ul>\n\n\n\n
      \u4ee3\u7801\u7247\u6bb5\uff1aagent\/component\/answer.py<\/strong><\/p>\n\n\n\n
      class AnswerComponent(BaseComponent):\n    def run(self, query):\n        # \u8fd9\u91cc\u53ef\u4ee5\u63a5 LLM\uff0c\u4e5f\u53ef\u4ee5\u63a5\u81ea\u5b9a\u4e49\u903b\u8f91\n        return self.llm.generate(query)<\/code><\/pre>\n\n\n\n
      \n
      \u201c\u6bcf\u4e2a\u667a\u80fd\u4f53\u7ec4\u4ef6\u90fd\u50cf\u662f\u4e00\u4e2a\u4e50\u9ad8\u79ef\u6728\uff0c\u968f\u65f6\u53ef\u4ee5\u62fc\u88c5\u51fa\u65b0\u73a9\u6cd5\u3002\u201d<\/p>\n<\/blockquote>\n\n\n\n
      2. agentic_reasoning\/ \u2014\u2014 \u591a\u667a\u80fd\u4f53\u63a8\u7406<\/h3>\n\n\n\n
      \u8fd9\u91cc\u662f\u201c\u7fa4\u4f53\u667a\u6167\u201d\u7684\u4f53\u73b0\u3002deep_research.py \u91cc\u5b9e\u73b0\u4e86\u591a\u667a\u80fd\u4f53\u534f\u4f5c\u63a8\u7406\uff0c\u6bd4\u5982\u8ba9\u591a\u4e2a agent \u5206\u5de5\u5408\u4f5c\uff0c\u4e92\u76f8\u8fa9\u8bba\uff0c\u6700\u540e\u7ed9\u51fa\u66f4\u9760\u8c31\u7684\u7b54\u6848\u3002<\/p>\n\n\n\n
      \u4ee3\u7801\u7247\u6bb5\uff1aagentic_reasoning\/deep_research.py<\/strong><\/p>\n\n\n\n
      def deep_research(query):\n    # \u591a\u667a\u80fd\u4f53\u5206\u5de5\u534f\u4f5c\n    results = [agent.run(query) for agent in agents]\n    return aggregate(results)<\/code><\/pre>\n\n\n\n
      \n
      \u201c\u4e00\u4e2a\u4eba\u662f\u667a\u6167\uff0c\u56e2\u961f\u662f\u529b\u91cf\u3002\u201d<\/p>\n<\/blockquote>\n\n\n\n
      3. api\/ \u2014\u2014 \u540e\u7aefAPI\u670d\u52a1<\/h3>\n\n\n\n
      \u8fd9\u91cc\u662f\u6574\u4e2a\u7cfb\u7edf\u7684\u201c\u4e2d\u67a2\u795e\u7ecf\u201d\u3002\u91c7\u7528 FastAPI\uff0c\u8d1f\u8d23\u5bf9\u5916\u63d0\u4f9b HTTP API\uff0c\u7ba1\u7406\u7528\u6237\u3001\u77e5\u8bc6\u5e93\u3001\u6570\u636e\u96c6\u3001\u4f1a\u8bdd\u3001\u6743\u9650\u7b49\u3002<\/p>\n\n\n\n
        \n
      • apps\/<\/strong>\uff1a\u6bcf\u4e2a\u5b50\u6a21\u5757\u4e00\u4e2a app\uff0c\u6bd4\u5982 auth\uff08\u8ba4\u8bc1\uff09\u3001canvas_app\uff08\u753b\u5e03\uff09\u3001sdk\uff08SDK API\uff09\u7b49\u3002<\/li>\n\n\n\n
      • db\/<\/strong>\uff1a\u6570\u636e\u5e93\u6a21\u578b\u4e0e\u670d\u52a1\uff0cSQLAlchemy ORM\u3002<\/li>\n\n\n\n
      • utils\/<\/strong>\uff1a\u901a\u7528\u5de5\u5177\u51fd\u6570\u3002<\/li>\n\n\n\n
      • ragflow_server.py<\/strong>\uff1aFastAPI \u542f\u52a8\u5165\u53e3\u3002<\/li>\n<\/ul>\n\n\n\n
        \u4ee3\u7801\u7247\u6bb5\uff1aapi\/apps\/api_app.py<\/strong><\/p>\n\n\n\n
        app = FastAPI()\n@app.get(\"\/api\/agent\/run\")\ndef run_agent(agent_id: str, input: dict):\n    # \u8c03\u7528 agent \u6267\u884c\n    return agent_manager.run(agent_id, input)<\/code><\/pre>\n\n\n\n
        \n
        \u201cAPI \u5c31\u50cf\u662f\u5927\u8111\u7684\u795e\u7ecf\u5143\uff0c\u8d1f\u8d23\u4fe1\u606f\u7684\u8f93\u5165\u8f93\u51fa\u3002\u201d<\/p>\n<\/blockquote>\n\n\n\n
        4. web\/ \u2014\u2014 \u524d\u7aef React \u5e94\u7528<\/h3>\n\n\n\n
        UI \u91c7\u7528 React + TypeScript\uff0cAnt Design \u7ec4\u4ef6\u5e93\uff0cUmiJS \u8def\u7531\u3002\u652f\u6301\u753b\u5e03\u5f0f agent \u6d41\u7a0b\u7f16\u6392\u3001\u77e5\u8bc6\u5e93\u7ba1\u7406\u3001\u804a\u5929\u3001\u56e2\u961f\u534f\u4f5c\u7b49\u529f\u80fd\u3002<\/p>\n\n\n\n
          \n
        • src\/components\/<\/strong>\uff1aUI \u7ec4\u4ef6\uff0c\u753b\u5e03\u3001\u804a\u5929\u3001\u77e5\u8bc6\u5e93\u7b49\u3002<\/li>\n\n\n\n
        • src\/pages\/<\/strong>\uff1a\u9875\u9762\u8def\u7531\u3002<\/li>\n\n\n\n
        • src\/services\/<\/strong>\uff1a\u524d\u7aef API \u5c01\u88c5\u3002<\/li>\n\n\n\n
        • src\/hooks\/<\/strong>\uff1a\u81ea\u5b9a\u4e49 hooks\uff0c\u72b6\u6001\u7ba1\u7406\u3002<\/li>\n<\/ul>\n\n\n\n
          \u4ee3\u7801\u7247\u6bb5\uff1aweb\/src\/components\/AgentCanvas.tsx<\/strong><\/p>\n\n\n\n
          const AgentCanvas = () => {\n  \/\/ \u753b\u5e03\u5f0f\u62d6\u62fd\u7f16\u6392\n  return <Canvas nodes={nodes} edges={edges} \/>;\n};<\/code><\/pre>\n\n\n\n
          \n
          \u201c\u524d\u7aef\u662f\u989c\u503c\u62c5\u5f53\uff0c\u8d1f\u8d23\u628a\u590d\u6742\u7684 AI \u80fd\u529b\u53d8\u6210\u4eba\u4eba\u53ef\u7528\u7684\u5de5\u5177\u3002\u201d<\/p>\n<\/blockquote>\n\n\n\n
          5. rag\/ \u2014\u2014 RAG \u6838\u5fc3\u4e0e NLP<\/h3>\n\n\n\n
          \u8fd9\u91cc\u662f\u201c\u77e5\u8bc6\u589e\u5f3a\u201d\u7684\u53d1\u52a8\u673a\u3002\u5b9e\u73b0\u4e86 RAG pipeline\u3001\u5206\u8bcd\u3001NER\u3001\u77e5\u8bc6\u5e93\u68c0\u7d22\u3001LLM \u8c03\u7528\u7b49\u3002<\/p>\n\n\n\n
            \n
          • llm\/<\/strong>\uff1a\u5927\u6a21\u578b\u9002\u914d\u5c42\uff0c\u652f\u6301\u591a\u79cd LLM\u3002<\/li>\n\n\n\n
          • nlp\/<\/strong>\uff1a\u5206\u8bcd\u3001NER\u3001tokenizer\u3002<\/li>\n\n\n\n
          • app\/<\/strong>\uff1aRAG \u5e94\u7528\u573a\u666f\uff0c\u5982\u97f3\u9891\u3001\u56fe\u4e66\u3001PDF\u3002<\/li>\n\n\n\n
          • utils\/<\/strong>\uff1aNLP \u5de5\u5177\u51fd\u6570\u3002<\/li>\n<\/ul>\n\n\n\n
            \u4ee3\u7801\u7247\u6bb5\uff1arag\/nlp\/query.py<\/strong><\/p>\n\n\n\n
            def split_query(query):\n    # \u5206\u8bcd+NER\n    tokens = tokenizer.tokenize(query)\n    entities = ner.extract(query)\n    return tokens, entities<\/code><\/pre>\n\n\n\n
            \n
            \u201cRAG \u8ba9\u5927\u6a21\u578b\u4e0d\u518d\u2018\u5065\u5fd8\u2019\uff0c\u77e5\u8bc6\u5e93\u968f\u67e5\u968f\u7528\u3002\u201d<\/p>\n<\/blockquote>\n\n\n\n
            6. plugin\/ \u2014\u2014 \u63d2\u4ef6\u673a\u5236<\/h3>\n\n\n\n
            \u652f\u6301 LLM \u5de5\u5177\u94fe\u6269\u5c55\uff0c\u6bd4\u5982\u5185\u7f6e\u7684 bad_calculator.py\uff08\u4e00\u4e2a\u201c\u7b97\u9519\u9898\u201d\u7684\u8ba1\u7b97\u5668\uff0c\u4e13\u6cbb\u5927\u6a21\u578b\u5e7b\u89c9\uff09\uff0c\u4e5f\u53ef\u4ee5\u81ea\u5b9a\u4e49\u63d2\u4ef6\u3002<\/p>\n\n\n\n
              \n
            • embedded_plugins\/<\/strong>\uff1a\u5185\u7f6e\u63d2\u4ef6\u3002<\/li>\n\n\n\n
            • llm_tool_plugin.py<\/strong>\uff1a\u63d2\u4ef6\u52a0\u8f7d\u4e0e\u6ce8\u518c\u3002<\/li>\n<\/ul>\n\n\n\n
              \u4ee3\u7801\u7247\u6bb5\uff1aplugin\/llm_tool_plugin.py<\/strong><\/p>\n\n\n\n
              def load_plugins():\n    # \u52a8\u6001\u52a0\u8f7d\u63d2\u4ef6\n    for plugin in plugin_dir:\n        register(plugin)<\/code><\/pre>\n\n\n\n
              \n
              \u201c\u63d2\u4ef6\u673a\u5236\u8ba9\u7cfb\u7edf\u50cf iPhone \u4e00\u6837\uff0c\u6c38\u8fdc\u6709\u65b0\u73a9\u6cd5\u3002\u201d<\/p>\n<\/blockquote>\n\n\n\n
              7. sandbox\/ \u2014\u2014 \u6c99\u7bb1\u4ee3\u7801\u6267\u884c<\/h3>\n\n\n\n
              \u5b89\u5168\u5730\u6267\u884c\u7528\u6237\u4ee3\u7801\uff0c\u652f\u6301 Python\/Nodejs\u3002\u91c7\u7528\u5bb9\u5668\u9694\u79bb\uff0c\u9632\u6b62\u201c\u718a\u5b69\u5b50\u201d\u628a\u670d\u52a1\u5668\u73a9\u574f\u3002<\/p>\n\n\n\n
                \n
              • executor_manager\/<\/strong>\uff1a\u6838\u5fc3\u670d\u52a1\uff0cAPI\u3001\u5bb9\u5668\u7ba1\u7406\u3001\u9650\u6d41\u3002<\/li>\n\n\n\n
              • sandbox_base_image\/<\/strong>\uff1a\u57fa\u7840\u955c\u50cf\uff0c\u652f\u6301\u591a\u8bed\u8a00\u3002<\/li>\n\n\n\n
              • scripts\/<\/strong>\uff1a\u542f\u52a8\/\u91cd\u542f\u811a\u672c\u3002<\/li>\n<\/ul>\n\n\n\n
                \u4ee3\u7801\u7247\u6bb5\uff1asandbox\/executor_manager\/services\/execution.py<\/strong><\/p>\n\n\n\n
                def execute_code(code, lang):\n    # \u5bb9\u5668\u9694\u79bb\u6267\u884c\n    container = create_container(lang)\n    result = container.run(code)\n    return result<\/code><\/pre>\n\n\n\n
                \n
                \u201c\u6c99\u7bb1\u662f\u2018\u6e38\u4e50\u573a\u2019\uff0c\u4f46\u6709\u62a4\u680f\uff0c\u5b89\u5168\u7b2c\u4e00\u3002\u201d<\/p>\n<\/blockquote>\n\n\n\n
                8. sdk\/ \u2014\u2014 Python SDK<\/h3>\n\n\n\n
                \u65b9\u4fbf\u5f00\u53d1\u8005\u4e8c\u6b21\u5f00\u53d1\uff0c\u5feb\u901f\u96c6\u6210 ragflow \u80fd\u529b\u3002\u652f\u6301\u6570\u636e\u96c6\u3001\u77e5\u8bc6\u5e93\u3001agent \u7b49 API \u5c01\u88c5\u3002<\/p>\n\n\n\n
                  \n
                • ragflow_sdk\/<\/strong>\uff1aSDK \u6838\u5fc3\u4ee3\u7801\u3002<\/li>\n\n\n\n
                • test\/<\/strong>\uff1aSDK \u6d4b\u8bd5\u7528\u4f8b\u3002<\/li>\n<\/ul>\n\n\n\n
                  \u4ee3\u7801\u7247\u6bb5\uff1asdk\/python\/ragflow_sdk\/ragflow.py<\/strong><\/p>\n\n\n\n
                  class RagflowClient:\n    def run_agent(self, agent_id, input):\n        # \u8c03\u7528\u540e\u7aef API\n        return requests.post(f\"{self.base_url}\/api\/agent\/run\", json=input)<\/code><\/pre>\n\n\n\n
                  \n
                  \u201cSDK \u5c31\u50cf\u662f\u2018\u9065\u63a7\u5668\u2019\uff0c\u8ba9\u4f60\u8f7b\u677e\u64cd\u63a7 ragflow\u3002\u201d<\/p>\n<\/blockquote>\n\n\n\n
                  9. deepdoc\/ \u2014\u2014 \u6587\u6863\u89e3\u6790\u4e0e OCR<\/h3>\n\n\n\n
                  \u652f\u6301 docx\u3001excel\u3001\u56fe\u7247\u7b49\u6587\u6863\u89e3\u6790\uff0c\u5185\u7f6e OCR \u80fd\u529b\uff0c\u65b9\u4fbf\u77e5\u8bc6\u5e93\u6784\u5efa\u3002<\/p>\n\n\n\n
                    \n
                  • parser\/<\/strong>\uff1a\u5404\u7c7b\u6587\u6863\u89e3\u6790\u5668\u3002<\/li>\n\n\n\n
                  • vision\/<\/strong>\uff1aOCR\u3001\u7248\u9762\u5206\u6790\u3002<\/li>\n<\/ul>\n\n\n\n
                    \u4ee3\u7801\u7247\u6bb5\uff1adeepdoc\/parser\/docx_parser.py<\/strong><\/p>\n\n\n\n
                    def parse_docx(file_path):\n    # \u89e3\u6790 docx \u6587\u6863\n    doc = Document(file_path)\n    return extract_text(doc)<\/code><\/pre>\n\n\n\n
                    \n
                    \u201cdeepdoc \u8ba9\u77e5\u8bc6\u5e93\u2018\u5403\u2019\u4e0b\u5404\u79cd\u6587\u6863\uff0c\u6d88\u5316\u6210\u77e5\u8bc6\u3002\u201d<\/p>\n<\/blockquote>\n\n\n\n
                    10. graphrag\/ \u2014\u2014 \u56fe\u8c31RAG<\/h3>\n\n\n\n
                    \u652f\u6301\u77e5\u8bc6\u56fe\u8c31\u62bd\u53d6\u3001\u5b9e\u4f53\u6d88\u6b67\u3001\u56fe\u8c31\u589e\u5f3a RAG\u3002<\/p>\n\n\n\n
                      \n
                    • entity_resolution.py<\/strong>\uff1a\u5b9e\u4f53\u6d88\u6b67\u3002<\/li>\n\n\n\n
                    • general\/<\/strong>\u3001light\/<\/strong>\uff1a\u4e0d\u540c\u56fe\u8c31\u62bd\u53d6\u7b56\u7565\u3002<\/li>\n<\/ul>\n\n\n\n
                      \u4ee3\u7801\u7247\u6bb5\uff1agraphrag\/entity_resolution.py<\/strong><\/p>\n\n\n\n
                      def resolve_entity(entity, context):\n    # \u56fe\u8c31\u6d88\u6b67\n    candidates = search_graph(entity)\n    return best_match(candidates, context)<\/code><\/pre>\n\n\n\n
                      \n
                      \u201cgraphrag \u8ba9 RAG \u4e0d\u6b62\u4e8e\u6587\u672c\uff0c\u77e5\u8bc6\u56fe\u8c31\u52a0\u6301\uff0c\u63a8\u7406\u66f4\u5f3a\u3002\u201d<\/p>\n<\/blockquote>\n\n\n\n
                      \n\n\n\n
                      \u56db\u3001\u90e8\u7f72\u4e0e\u8fd0\u7ef4<\/h2>\n\n\n\n
                        \n
                      • Dockerfile\/Helm<\/strong>\uff1a\u4e00\u952e\u90e8\u7f72\uff0c\u652f\u6301\u4e91\u539f\u751f\u3002<\/li>\n\n\n\n
                      • conf\/<\/strong>\uff1a\u591a\u73af\u5883\u914d\u7f6e\uff0c\u652f\u6301 LLM\u3001\u63d2\u4ef6\u3001\u77e5\u8bc6\u5e93\u7b49\u7075\u6d3b\u5207\u6362\u3002<\/li>\n\n\n\n
                      • start_backend.sh\/start_frontend.sh<\/strong>\uff1a\u672c\u5730\u5f00\u53d1\u4e00\u952e\u542f\u52a8\u3002<\/li>\n<\/ul>\n\n\n\n
                        \n\n\n\n
                        \u4e94\u3001\u603b\u7ed3<\/h2>\n\n\n\n
                        ragflow-main \u5c31\u50cf\u662f\u201cAI \u4e50\u9ad8\u79ef\u6728\u201d\uff0c\u6bcf\u4e2a\u6a21\u5757\u90fd\u662f\u4e00\u4e2a\u529f\u80fd\u5757\uff0c\u65e2\u80fd\u5355\u72ec\u73a9\uff0c\u4e5f\u80fd\u62fc\u6210\u8d85\u7ea7\u7cfb\u7edf\u3002\u65e0\u8bba\u4f60\u662f\u60f3\u505a\u667a\u80fd\u95ee\u7b54\u3001\u77e5\u8bc6\u5e93\u3001RAG\u3001\u8fd8\u662f\u591a\u667a\u80fd\u4f53\u534f\u4f5c\uff0c\u8fd9\u91cc\u90fd\u80fd\u627e\u5230\u73b0\u6210\u7684\u201c\u79ef\u6728\u201d\u3002<\/p>\n\n\n\n
                        \n
                        \u201c\u67b6\u6784\u8981\u4f18\u96c5\uff0c\u4ee3\u7801\u8981\u6709\u8da3\uff0cAI \u8981\u806a\u660e\uff0c\u5e73\u53f0\u8981\u5f00\u653e\u3002\u201d<\/p>\n<\/blockquote>\n\n\n\n
                        \u5982\u679c\u4f60\u60f3\u6df1\u5165\u4f53\u9a8c ragflow\uff0c\u4e0d\u59a8 clone \u4e0b\u6765\uff0c\u8dd1\u4e00\u8dd1\uff0c\u62fc\u4e00\u62fc\uff0c\u8bf4\u4e0d\u5b9a\u4e0b\u4e00\u4e2a\u201cAI \u7206\u6b3e\u201d\u5c31\u51fa\u81ea\u4f60\u624b\uff01<\/p>\n\n\n\n
                        \n\n\n\n
                        \u5982\u9700\u66f4\u6df1\u5165\u7684\u4ee3\u7801\u5206\u6790\u6216\u67d0\u4e2a\u6a21\u5757\u7684\u8be6\u7ec6\u89e3\u8bfb\uff0c\u6b22\u8fce\u7559\u8a00\u201c\u70b9\u83dc\u201d\uff01<\/p>\n","protected":false},"excerpt":{"rendered":"
                        \u201c\u4ee3\u7801\u5982\u8bd7\uff0c\u67b6\u6784\u5982\u753b\u3002\u4eca\u5929\u6211\u4eec\u6765...<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[74],"tags":[73],"class_list":["post-2546","post","type-post","status-publish","format-standard","hentry","category-rag","tag-ai"],"_links":{"self":[{"href":"https:\/\/sanlangcode.com\/index.php\/wp-json\/wp\/v2\/posts\/2546","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/sanlangcode.com\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/sanlangcode.com\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/sanlangcode.com\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/sanlangcode.com\/index.php\/wp-json\/wp\/v2\/comments?post=2546"}],"version-history":[{"count":1,"href":"https:\/\/sanlangcode.com\/index.php\/wp-json\/wp\/v2\/posts\/2546\/revisions"}],"predecessor-version":[{"id":2547,"href":"https:\/\/sanlangcode.com\/index.php\/wp-json\/wp\/v2\/posts\/2546\/revisions\/2547"}],"wp:attachment":[{"href":"https:\/\/sanlangcode.com\/index.php\/wp-json\/wp\/v2\/media?parent=2546"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/sanlangcode.com\/index.php\/wp-json\/wp\/v2\/categories?post=2546"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/sanlangcode.com\/index.php\/wp-json\/wp\/v2\/tags?post=2546"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}