{"id":1157,"date":"2025-07-28T16:30:00","date_gmt":"2025-07-28T08:30:00","guid":{"rendered":"http:\/\/www.applicationsofllm.com\/?p=1157"},"modified":"2025-09-17T16:41:23","modified_gmt":"2025-09-17T08:41:23","slug":"deepseek%e3%80%81gpt-5%e9%83%bd%e5%9c%a8%e5%b0%9d%e8%af%95%e7%9a%84%e5%bf%ab%e6%85%a2%e6%80%9d%e8%80%83%e5%88%87%e6%8d%a2%ef%bc%8c%e6%9c%89%e4%ba%86%e6%9b%b4%e6%99%ba%e8%83%bd%e7%89%88%e6%9c%ac","status":"publish","type":"post","link":"http:\/\/www.applicationsofllm.com\/index.php\/2025\/07\/28\/deepseek%e3%80%81gpt-5%e9%83%bd%e5%9c%a8%e5%b0%9d%e8%af%95%e7%9a%84%e5%bf%ab%e6%85%a2%e6%80%9d%e8%80%83%e5%88%87%e6%8d%a2%ef%bc%8c%e6%9c%89%e4%ba%86%e6%9b%b4%e6%99%ba%e8%83%bd%e7%89%88%e6%9c%ac\/","title":{"rendered":"DeepSeek\u3001GPT-5\u90fd\u5728\u5c1d\u8bd5\u7684\u5feb\u6162\u601d\u8003\u5207\u6362\uff0c\u6709\u4e86\u66f4\u667a\u80fd\u7248\u672c\uff0c\u8fd8\u662f\u591a\u6a21\u6001"},"content":{"rendered":"\n

\u80cc\u666f\uff1a\u591a\u6a21\u6001\u5927\u6a21\u578b\u7684\u601d\u8003\u56f0\u5883<\/strong><\/p>\n\n\n\n

\u5f53\u524d\uff0c\u4e1a\u754c\u9876\u5c16\u7684\u5927\u6a21\u578b\u6b63\u7ade\u76f8\u6311\u6218\u201c\u8fc7\u5ea6\u601d\u8003<\/strong>\u201d\u7684\u96be\u9898\uff0c\u5373\u65e0\u8bba\u95ee\u9898\u7b80\u5355\u4e0e\u5426\uff0c\u5b83\u4eec\u90fd\u91c7\u7528 \u201calways-on thinking\u201d \u7684\u8be6\u7ec6\u63a8\u7406\u6a21\u5f0f\u3002\u65e0\u8bba\u662f\u50cf DeepSeek-V3.1 \u8fd9\u79cd\u4f9d\u8d56\u6df7\u5408\u63a8\u7406\u67b6\u6784\u63d0\u4f9b\u9700\u7528\u6237\u201c\u624b\u52a8\u201d\u4ecb\u5165\u7684\u5feb\u6162\u601d\u8003\u5207\u6362\uff0c\u8fd8\u662f\u5982 GPT-5 \u90a3\u6837\u901a\u8fc7\u4f9d\u8d56\u5e9e\u5927\u800c\u9ad8\u6210\u672c\u7684\u201c\u4e13\u5bb6\u8def\u7531\u201d\u673a\u5236\u63d0\u4f9b\u7684\u81ea\u9002\u5e94\u601d\u8003\u5207\u6362\u3002\u5b83\u4eec\u8ddd\u79bb\u771f\u6b63\u610f\u4e49\u4e0a\u7684\u201c\u667a\u80fd\u601d\u8003\u201d<\/strong>\u4ecd\u6709\u8ddd\u79bb\u3002\u8fd9\u4e9b\u65b9\u6848\u6216\u5c06\u5224\u65ad\u538b\u529b\u8f6c\u79fb\u7ed9\u7528\u6237\uff0c\u6216\u53d7\u9650\u4e8e\u590d\u6742\u7684\u7cfb\u7edf\u67b6\u6784\u548c\u9ad8\u6602\u7684\u90e8\u7f72\u6210\u672c\u3002\u56e0\u6b64\uff0c\u7814\u53d1\u4e00\u6b3e\u8f7b\u91cf\u5316\u3001\u652f\u6301\u591a\u6a21\u6001\u4e14\u80fd\u5b9e\u73b0\u66f4\u667a\u80fd\u81ea\u9002\u5e94\u601d\u8003\u7684\u5927\u6a21\u578b<\/strong>\uff0c\u5c06\u4e3a\u7528\u6237\u63d0\u4f9b\u66f4\u52a0\u6d41\u7545\u7684\u4ea4\u4e92\u4f53\u9a8c\u3002<\/p>\n\n\n\n

\u8fd1\u671f\uff0c\u7531\u817e\u8baf\u6df7\u5143\u56e2\u961f\u4e0e\u4e2d\u79d1\u9662\u81ea\u52a8\u5316\u6240\u5408\u4f5c\u7684\u4e00\u9879\u6700\u65b0\u7814\u7a76\u63a8\u51fa R-4B \u591a\u6a21\u6001\u5927\u6a21\u578b\uff0c\u901a\u8fc7\u81ea\u9002\u5e94\u601d\u8003\uff08auto-thinking\uff09<\/strong>\u673a\u5236\uff0c\u6539\u53d8\u4e86\u8fd9\u4e00\u73b0\u72b6\uff0c\u5b83\u8ba9 AI \u80fd\u50cf\u4eba\u7c7b\u4e00\u6837 \u201c\u667a\u80fd\u5207\u6362\u201d \u601d\u7ef4\u6a21\u5f0f\u3002\u7b80\u5355\u95ee\u9898\u76f4\u63a5\u54cd\u5e94\uff0c\u590d\u6742\u95ee\u9898\u6df1\u5ea6\u63a8\u7406\uff0c\u5728\u6700\u5927\u5316\u56de\u7b54\u51c6\u786e\u6027\u7684\u540c\u65f6\uff0c\u6700\u5c0f\u5316\u8ba1\u7b97\u5f00\u9500\u3002<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n
    \n
  • \u8bba\u6587\u6807\u9898\uff1aR-4B: INCENTIVIZING GENERAL-PURPOSE AUTOTHINKING CAPABILITY IN MLLMS VIA BI-MODE ANNEALING AND REINFORCE LEARNING<\/li>\n\n\n\n
  • \u8bba\u6587\u94fe\u63a5\uff1ahttps:\/\/arxiv.org\/pdf\/2508.21113<\/li>\n<\/ul>\n\n\n\n

    \u8fd9\u4e00 \u201c\u6309\u9700\u601d\u8003\u201d \u7684\u6838\u5fc3\u80fd\u529b\uff0c\u4e3a 4B \u91cf\u7ea7\u7684\u591a\u6a21\u6001\u6a21\u578b\u6811\u7acb\u4e86\u5168\u65b0\u7684\u6027\u80fd\u6807\u6746\uff0c\u4f7f\u5176\u5728\u8bc4\u6d4b\u6027\u80fd\u6307\u6807\u4e0a\u6210\u529f\u8d85\u8d8a\u4e86 Keye-VL-8B\u3001Kimi-VL-A3B-Thinking-2506 \u7b49\u66f4\u5927\u89c4\u6a21\u7684\u6a21\u578b\u3002<\/strong><\/p>\n\n\n\n

    \"\"<\/figure>\n\n\n\n

    \u540c\u65f6\uff0cR-4B \u5728\u6743\u5a01\u57fa\u51c6 OpenCompass \u699c\u5355\u4e0a\u53d6\u5f97\u4e86\u4f18\u5f02\u6210\u7ee9\u3002<\/strong><\/p>\n\n\n\n

      \n
    • \u767b\u9876 OpenCompass \u591a\u6a21\u6001\u5b66\u672f\u699c\u5355\uff1a\u5728 20B \u4ee5\u5185\u89c4\u6a21\u591a\u6a21\u6001\u5927\u6a21\u578b\u4e2d\uff0c\u6027\u80fd\u6392\u540d Top 1\uff01<\/li>\n<\/ul>\n\n\n\n
      \"\"<\/figure>\n\n\n\n
        \n
      • \u4f4d\u5217 OpenCompass \u591a\u6a21\u6001\u63a8\u7406\u699c\u5355\u5f00\u6e90\u699c\u9996\uff1a\u5728\u5f00\u6e90\u6a21\u578b\u4e2d\uff0c\u63a8\u7406\u6027\u80fd\u62d4\u5f97\u5934\u7b79\uff01\u00a0<\/li>\n<\/ul>\n\n\n\n
        \"\"<\/figure>\n\n\n\n

        \u76ee\u524d\uff0c\u8be5\u6a21\u578b\u5df2\u5728 GitHub \u548c HuggingFace \u4e0a\u7ebf\uff0c\u4e14\u652f\u6301 vLLM \u5feb\u901f\u90e8\u7f72\u3002\u300c\u6d88\u8d39\u7ea7\u663e\u5361\u5373\u53ef\u8fd0\u884c\uff0c\u9002\u7528\u4e8e\u7b14\u8bb0\u672c\u7535\u8111\u3001\u667a\u80fd\u5ea7\u8231\u3001\u667a\u80fd\u5bb6\u5c45\u7b49\u4f4e\u529f\u8017\u573a\u666f\uff0c\u652f\u6301\u5782\u76f4\u9886\u57df\u4f4e\u6210\u672c\u5fae\u8c03\u3002\u300d\u622a\u81f3\u76ee\u524d\u4e0b\u8f7d\u91cf\u5df2\u7834\u4e07\uff0c\u6b22\u8fce\u5927\u5bb6\u4f53\u9a8c\uff01<\/strong><\/p>\n\n\n\n

          \n
        • GitHub \u4ee3\u7801\u4ed3\u5e93\uff1ahttps:\/\/github.com\/yannqi\/R-4B\u00a0<\/li>\n\n\n\n
        • Hugging Face \u6a21\u578b\u4e0b\u8f7d\uff1ahttps:\/\/huggingface.co\/YannQi\/R-4B<\/li>\n<\/ul>\n\n\n\n

          \u7a81\u7834\uff1aR-4B \u7684\u81ea\u9002\u5e94\u601d\u8003\u5f15\u64ce<\/strong><\/p>\n\n\n\n

          R-4B \u7684\u667a\u6167\u4e4b\u5904\u5728\u4e8e\u5176\u81ea\u9002\u5e94\u601d\u8003\u80fd\u529b\uff1a<\/strong><\/p>\n\n\n\n

          \"\"<\/figure>\n\n\n\n
            \n
          • \u9047\u5230\u7b80\u5355\u95ee\u9898\uff08\u7b80\u5355\u5b9e\u4f53\u8bc6\u522b\u3001\u7b80\u6613\u95ee\u7b54\uff09\uff0c\u5b83\u9009\u62e9\u76f4\u63a5\u3001\u9ad8\u6548\u5730\u54cd\u5e94\u3002<\/li>\n\n\n\n
          • \u9762\u5bf9\u590d\u6742\u4efb\u52a1\uff08\u5982\u6570\u5b66\u8ba1\u7b97\u3001\u56fe\u8868\u5206\u6790\uff09\uff0c\u5b83\u5219\u81ea\u52a8\u5207\u6362\u5230\u6df1\u5ea6\u601d\u8003\u6a21\u5f0f\uff0c\u751f\u6210\u8be6\u7ec6\u7684\u601d\u8003\u8fc7\u7a0b\u3002<\/li>\n<\/ul>\n\n\n\n

            R-4B \u7684\u6838\u5fc3\u521b\u65b0\u5728\u4e8e\u5176\u72ec\u7279\u7684\u4e24\u9636\u6bb5\u8bad\u7ec3\u7b56\u7565\u3002\u4e3a\u5b9e\u73b0\u6a21\u578b\u5728\u901a\u7528\u9886\u57df\u7684\u81ea\u9002\u5e94\u601d\u8003\uff0c\u7814\u7a76\u56e2\u961f\u9996\u5148\u63d0\u51fa\u53cc\u6a21\u9000\u706b<\/strong>\uff08bi-mode annealing\uff09\u8bad\u7ec3\u7b56\u7565\uff0c\u4fc3\u4f7f\u6a21\u578b\u540c\u65f6\u638c\u63e1\u901a\u7528\u9886\u57df\u7684\u601d\u8003\u4e0e\u975e\u601d\u8003\u80fd\u529b<\/strong>\u3002<\/p>\n\n\n\n

            \u8be5\u9636\u6bb5\u53ef\u4ee5\u7406\u89e3\u4e3a\u5bf9\u6a21\u578b\u8fdb\u884c \u201c\u601d\u8003\u201d \u542f\u8499\uff0c\u5373\u540c\u65f6\u5582\u7ed9\u5b83\u4e24\u79cd\u8303\u5f0f\u6570\u636e\uff1a\u4e00\u79cd\u9700\u8981\u76f4\u63a5\u56de\u7b54\uff08\u975e\u601d\u8003\u6a21\u5f0f\uff0c\u50cf\u65e5\u5e38\u5bf9\u8bdd\uff09\uff0c\u53e6\u4e00\u79cd\u9700\u8981\u8be6\u7ec6\u63a8\u7406\uff08\u601d\u8003\u6a21\u5f0f\uff0c\u50cf\u89e3\u6570\u5b66\u9898\uff09\u3002\u901a\u8fc7\u8fd9\u79cd\u8bad\u7ec3\uff0c\u6a21\u578b\u540c\u65f6\u638c\u63e1\u4e86\u601d\u8003\u548c\u975e\u601d\u8003\u8fd9\u4e24\u79cd\u54cd\u5e94\u6a21\u5f0f\uff0c\u4e3a\u540e\u7eed\u7684\u81ea\u9002\u5e94\u601d\u8003\u6a21\u5f0f\u8bad\u7ec3\u6253\u4e0b\u575a\u5b9e\u57fa\u7840\u3002\u8be5\u9636\u6bb5\u7684\u6838\u5fc3\u662f\u901a\u7528\u9886\u57df\u63a8\u7406\u548c\u975e\u63a8\u7406\u6a21\u5f0f\u7684\u6570\u636e\u6784\u5efa\u7b56\u7565\uff1a\u9488\u5bf9\u5ba2\u89c2\u9898\uff0c\u7528\u6a21\u578b\u91c7\u6837\u7684\u7b54\u6848\u4e00\u81f4\u6027\u6765\u8861\u91cf\u9898\u76ee\u7684\u96be\u6613\u7a0b\u5ea6\uff1b\u9488\u5bf9\u4e3b\u89c2\u9898\u76ee\uff0c\u7528\u63d0\u793a\u5de5\u7a0b\u7684\u65b9\u5f0f\u53bb\u533a\u5206\u89e3\u51b3\u95ee\u9898\u662f\u5426\u9700\u8981\u8fdb\u4e00\u6b65\u601d\u8003\u3002<\/p>\n\n\n\n

            \"\"<\/figure>\n\n\n\n
              \n
            • \u63a8\u7406\u6a21\u5f0f\u6570\u636e\uff1a\u6db5\u76d6\u56fe\u8868\u5206\u6790\u3001\u903b\u8f91\u63a8\u7406\u7b49\u9700\u591a\u6b65\u63a8\u7406\u7684\u4efb\u52a1\uff08\u5982\u79d1\u5b66\u56fe\u89e3\u6216\u6570\u5b66\u95ee\u9898\uff09\u3002<\/li>\n\n\n\n
            • \u975e\u63a8\u7406\u6a21\u5f0f\u6570\u636e\uff1a\u9488\u5bf9\u76f4\u63a5\u4e8b\u5b9e\u54cd\u5e94\u7684\u67e5\u8be2\uff08\u5982\u5b9e\u4f53\u8bc6\u522b\u6216\u7b80\u5355\u95ee\u7b54\uff09\u3002<\/li>\n<\/ul>\n\n\n\n
              \"\"<\/figure>\n\n\n\n

              \u7ecf\u8fc7\u9000\u706b\u8bad\u7ec3\uff0c\u5f97\u5230\u4e00\u4e2a\u540c\u65f6\u7cbe\u901a\u601d\u8003\u4e0e\u975e\u601d\u8003\u6a21\u5f0f\u7684\u57fa\u7840\u6a21\u578b R-4B-Base \uff0c\u4e3a\u540e\u7eed\u81ea\u9002\u5e94\u601d\u8003\u5f3a\u5316\u8bad\u7ec3\u5960\u5b9a\u57fa\u7840\u3002\u57fa\u4e8e\u6b64\uff0c\u56e2\u961f\u5f00\u53d1\u4e86\u53cc\u6a21\u7b56\u7565\u4f18\u5316<\/strong>\uff08Bi-mode Policy Optimization, BPO\uff09\u5f3a\u5316\u5b66\u4e60\u7b97\u6cd5\u3002\u5b83\u65e0\u9700\u4f9d\u8d56\u7cbe\u5fc3\u8bbe\u8ba1\u7684\u5956\u52b1\u51fd\u6570\u6216\u7279\u5b9a\u6570\u636e\uff0c\u800c\u662f\u4ec5\u4f9d\u8d56\u57fa\u4e8e\u89c4\u5219\u7684\u5956\u52b1\u4fe1\u53f7\uff0c\u4ece\u6570\u5b66\u6570\u636e\u51fa\u53d1\uff0c\u5e76\u53ef\u6cdb\u5316\u5230\u901a\u7528\u9886\u57df\u3002\u5176\u6838\u5fc3\u662f\u6df7\u5408\u53cc\u6a21 rollout \u673a\u5236<\/strong>\uff0c\u901a\u8fc7\u5f3a\u5236\u6a21\u578b\u5728\u8bad\u7ec3\u4e2d\u540c\u65f6\u63a2\u7d22\u601d\u8003\u6a21\u5f0f\u548c\u975e\u601d\u8003\u6a21\u5f0f\u8f68\u8ff9\uff0c\u4ece\u800c\u907f\u514d\u6a21\u578b\u9677\u5165\u5bf9\u5355\u4e00\u6a21\u5f0f\u7684\u54cd\u5e94\u504f\u597d\u3002\u5728\u6b64\u57fa\u7840\u4e0a\uff0c\u901a\u8fc7\u540c\u65f6\u5956\u52b1\u4e24\u79cd\u601d\u8003\u6a21\u5f0f\u7684\u7b56\u7565<\/strong>\uff0c\u4f7f\u6a21\u578b\u81ea\u5df1\u5b66\u4f1a\u5224\u522b\u4f55\u65f6\u5e94\u8be5\u601d\u8003\u3002<\/p>\n\n\n\n

              \"\"<\/figure>\n\n\n\n

              \u6027\u80fd\u8868\u73b0\uff1a\u5c0f\u6a21\u578b\uff0c\u5927\u80fd\u91cf<\/strong><\/p>\n\n\n\n

              R-4B-RL \u6a21\u578b\u5728\u591a\u9879\u516c\u5f00\u57fa\u51c6\u6d4b\u8bd5\u4e2d\u6027\u80fd\u8868\u73b0\u5353\u8d8a\uff0c\u5237\u65b0\u4e86\u73b0\u6709\u8bb0\u5f55\uff0c\u5176\u6027\u80fd\u8d85\u8fc7 Keye-VL-8B\u3001Kimi-VL-A3B-Thinking-2506 \u7b49\u66f4\u5927\u89c4\u6a21\u7684\u6a21\u578b\u3002<\/strong><\/p>\n\n\n\n

              \"\"<\/figure>\n\n\n\n

              \u66f4\u5173\u952e\u7684\u662f\uff0cR-4B-RL \u5728\u81ea\u9002\u5e94\u601d\u8003\u6a21\u5f0f\u4e0b\u5b9e\u73b0\u4e86\u63a8\u7406\u6548\u7387\u7684\u63d0\u5347\uff0c\u5728\u7b80\u5355\u4efb\u52a1\u4e0b\u6a21\u578b\u65e0\u9700\u6d88\u8017\u66f4\u591a\u7684 Token<\/strong>\u3002\u8fd9\u8bc1\u660e\u4e86 BPO \u7b97\u6cd5\u7684\u6709\u6548\u6027\uff0c\u5373\u65e0\u9700\u901a\u7528\u9886\u57df\u7684\u5f3a\u5316\u5b66\u4e60\u6570\u636e\u6216\u989d\u5916\u7684\u5956\u52b1\u51fd\u6570\u8bbe\u8ba1\uff0c\u6a21\u578b\u4e5f\u80fd\u5b9e\u73b0\u81ea\u9002\u5e94\u601d\u8003\u3002<\/strong><\/p>\n\n\n\n

              \"\"<\/figure>\n\n\n\n

              \u5e94\u7528\u524d\u666f\uff1a\u4ece\u79d1\u7814\u5230\u4ea7\u4e1a\u7684\u667a\u80fd\u5316\u6d6a\u6f6e<\/strong><\/p>\n\n\n\n

              R-4B \u7684\u7a81\u7834\u4e0d\u6b62\u4e8e\u6280\u672f\uff0c\u66f4\u5f00\u542f\u4e86\u5e7f\u9614\u5e94\u7528\u573a\u666f\uff1a<\/p>\n\n\n\n

                \n
              • \u5e94\u7528\u667a\u80fd<\/strong>\u00a0\uff1a\u5728\u65e5\u5e38\u95ee\u7b54\u5206\u6790\u4e2d\uff0c\u81ea\u52a8\u5207\u6362\u7b80\u5355\u67e5\u8be2\uff08\u5982\u6587\u6863\u5185\u5bb9\u63d0\u53d6\uff09\u548c\u590d\u6742\u63a8\u7406\uff08\u5982\u56fe\u8868\u5206\u6790\uff09\u7684\u601d\u7ef4\u6a21\u5f0f\uff0c\u63d0\u5347\u81ea\u52a8\u5316\u5904\u7406\u6548\u7387\u3002<\/li>\n\n\n\n
              • \u79d1\u5b66\u7814\u7a76<\/strong>\u00a0\uff1a\u5728\u5904\u7406\u79d1\u5b66\u56fe\u8868\u65f6\uff0cR-4B \u7684\u6df1\u5ea6\u63a8\u7406\u6a21\u5f0f\u53ef\u89e3\u6790\u591a\u6b65\u5173\u7cfb\uff0c\u7cbe\u51c6\u89e3\u8bfb\u6570\u636e\uff0c\u63d0\u9ad8\u7814\u7a76\u6548\u7387\u3002<\/li>\n\n\n\n
              • \u6d88\u8d39\u7ea7 AI\u00a0<\/strong>\uff1a\u8fb9\u7f18\u8bbe\u5907\u90e8\u7f72\u4e2d\uff0cR-4B \u51ed\u501f\u66f4\u5c11\u7684\u53c2\u6570\u548c\u81ea\u9002\u5e94\u601d\u8003\u6a21\u5f0f\u964d\u4f4e\u5ef6\u8fdf\u548c\u80fd\u8017\uff0c\u9002\u7528\u4e8e\u5373\u65f6\u95ee\u7b54\u7cfb\u7edf\u3002<\/li>\n<\/ul>\n\n\n\n

                (1) \u6587\u6863\u5185\u5bb9\u63d0\u53d6\uff08\u7b80\u5355\u67e5\u8be2\uff09<\/p>\n\n\n\n

                \"\"<\/figure>\n\n\n\n

                (2) \u56fe\u8868\u5206\u6790\uff08\u590d\u6742\u63a8\u7406\uff09<\/p>\n\n\n\n

                \"\"<\/figure>\n\n\n\n

                \u7ed3\u8bed\uff1a\u81ea\u9002\u5e94\u601d\u8003\uff0c\u63a2\u7d22 AI \u53d1\u5c55\u65b0\u9053\u8def<\/strong><\/p>\n\n\n\n

                \u4ece\u53cc\u6a21\u9000\u706b\u8bad\u7ec3\u5230 BPO \u4f18\u5316\uff0cR-4B \u4e0d\u4ec5\u89e3\u51b3\u4e86 MLLMs \u7684\u601d\u8003\u56f0\u5883\uff0c\u66f4\u5728\u5c0f\u5c3a\u5bf8\u6a21\u578b\u4e0a\u63a2\u7d22\u4e86\u81ea\u9002\u5e94\u601d\u8003\u7684\u53ef\u884c\u6027 \u3002\u81ea\u9002\u5e94\u601d\u8003\u4e0d\u4ec5\u662f\u6280\u672f\u4f18\u5316\uff0c\u66f4\u662f\u5bf9\u6548\u7387\u4e0e\u666e\u60e0\u5e73\u8861\u7684\u8ffd\u6c42\u3002\u5728 AI \u8ba1\u7b97\u4e0e\u63a8\u7406\u6210\u672c\u98d9\u5347\u7684\u4eca\u5929\uff0cR-4B \u7684\u8f7b\u91cf\u5316\u3001\u667a\u80fd\u5316\u8bbe\u8ba1\uff0c\u4e3a\u5927\u6a21\u578b\u53ef\u6301\u7eed\u53d1\u5c55\u6ce8\u5165\u7eff\u8272\u52a8\u529b\u3002<\/p>\n","protected":false},"excerpt":{"rendered":"

                \u80cc\u666f\uff1a\u591a\u6a21\u6001\u5927\u6a21\u578b\u7684\u601d\u8003\u56f0\u5883 \u5f53\u524d\uff0c\u4e1a\u754c\u9876\u5c16\u7684\u5927\u6a21\u578b\u6b63\u7ade\u76f8\u6311\u6218\u201c\u8fc7\u5ea6\u601d\u8003\u201d\u7684\u96be\u9898\uff0c\u5373\u65e0\u8bba\u95ee\u9898\u7b80\u5355\u4e0e\u5426\uff0c\u5b83\u4eec\u90fd\u91c7 […]<\/p>\n","protected":false},"author":1,"featured_media":1159,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[10],"tags":[],"class_list":["post-1157","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-10"],"_links":{"self":[{"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/posts\/1157","targetHints":{"allow":["GET"]}}],"collection":[{"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/comments?post=1157"}],"version-history":[{"count":1,"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/posts\/1157\/revisions"}],"predecessor-version":[{"id":1171,"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/posts\/1157\/revisions\/1171"}],"wp:featuredmedia":[{"embeddable":true,"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/media\/1159"}],"wp:attachment":[{"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/media?parent=1157"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/categories?post=1157"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/www.applicationsofllm.com\/index.php\/wp-json\/wp\/v2\/tags?post=1157"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}