{"id":14910,"date":"2024-12-04T11:06:01","date_gmt":"2024-12-04T03:06:01","guid":{"rendered":"https:\/\/www.aisharenet.com\/?p=14910"},"modified":"2024-12-08T14:57:23","modified_gmt":"2024-12-08T06:57:23","slug":"aigongchengxueyuan24","status":"publish","type":"post","link":"https:\/\/www.kdjingpai.com\/ja\/aigongchengxueyuan24\/","title":{"rendered":"AI\u5de5\u7a0b\u5b66\u9662\uff1a2.4\u7528\u4e8e\u68c0\u7d22\u589e\u5f3a\u751f\u6210\uff08RAG\uff09\u7cfb\u7edf\u7684\u6570\u636e\u5206\u5757\u6280\u672f"},"content":{"rendered":"<h2>\u7b80\u4ecb<\/h2>\n<p>\u6570\u636e\u5206\u5757\u662f\u68c0\u7d22\u589e\u5f3a\u751f\u6210\uff08RAG\uff09\u7cfb\u7edf\u4e2d\u7684\u5173\u952e\u6b65\u9aa4\u3002\u5b83\u5c06\u5927\u578b\u6587\u6863\u5206\u89e3\u4e3a\u8f83\u5c0f\u7684\u3001\u53ef\u7ba1\u7406\u7684\u7247\u6bb5\uff0c\u4ee5\u4fbf\u9ad8\u6548\u7d22\u5f15\u3001\u68c0\u7d22\u548c\u5904\u7406\u3002\u672c README \u63d0\u4f9b\u4e86 <a href=\"https:\/\/www.kdjingpai.com\/de\/rag\/\">RAG<\/a> \u7ba1\u9053\u4e2d\u53ef\u7528\u7684\u5404\u79cd\u5206\u5757\u65b9\u6cd5\u6982\u8ff0\u3002<\/p>\n<p>https:\/\/github.com\/adithya-s-k\/AI-Engineering.academy\/tree\/main\/RAG\/01_Data_Ingestion<\/p>\n<p>&nbsp;<\/p>\n<h2>\u5206\u5757\u5728 RAG \u4e2d\u7684\u91cd\u8981\u6027<\/h2>\n<p>\u6709\u6548\u7684\u5206\u5757\u5bf9 RAG \u7cfb\u7edf\u81f3\u5173\u91cd\u8981\uff0c\u56e0\u4e3a\u5b83\u53ef\u4ee5\uff1a<\/p>\n<ol>\n<li>\u901a\u8fc7\u521b\u5efa\u8fde\u8d2f\u7684\u81ea\u5305\u542b\u4fe1\u606f\u5355\u5143\uff0c\u63d0\u9ad8\u68c0\u7d22\u51c6\u786e\u6027\u3002<\/li>\n<li>\u63d0\u5347\u5d4c\u5165\u751f\u6210\u4e0e\u76f8\u4f3c\u6027\u641c\u7d22\u7684\u6548\u7387\u3002<\/li>\n<li>\u5728\u751f\u6210\u54cd\u5e94\u65f6\uff0c\u5141\u8bb8\u66f4\u7cbe\u786e\u5730\u9009\u62e9\u4e0a\u4e0b\u6587\u3002<\/li>\n<li>\u5e2e\u52a9\u7ba1\u7406\u8bed\u8a00\u6a21\u578b\u548c\u5d4c\u5165\u7cfb\u7edf\u7684 <a href=\"https:\/\/www.kdjingpai.com\/de\/tokenization\/\">Token<\/a> \u9650\u5236\u3002<\/li>\n<\/ol>\n<p>&nbsp;<\/p>\n<h2>\u5206\u5757\u65b9\u6cd5<\/h2>\n<p>\u6211\u4eec\u5df2\u5b9e\u73b0\u4e86\u516d\u79cd\u4e0d\u540c\u7684\u5206\u5757\u65b9\u6cd5\uff0c\u6bcf\u79cd\u65b9\u6cd5\u5177\u6709\u4e0d\u540c\u7684\u4f18\u52bf\u548c\u4f7f\u7528\u573a\u666f\uff1a<\/p>\n<ol>\n<li>RecursiveCharacterTextSplitter<\/li>\n<li>TokenTextSplitter<\/li>\n<li>KamradtSemanticChunker<\/li>\n<li>KamradtModifiedChunker<\/li>\n<li>ClusterSemanticChunker<\/li>\n<li>LLMSemanticChunker<\/li>\n<\/ol>\n<p>&nbsp;<\/p>\n<h2>\u5206\u5757\u5de5\u4f5c\u6d41<\/h2>\n<h3>1. RecursiveCharacterTextSplitter<\/h3>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"\" src=\"https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2024\/12\/83646a76f61d623.jpg\" alt=\"\" width=\"203\" height=\"483\" \/><\/p>\n<h3>2. TokenTextSplitter<\/h3>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"\" src=\"https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2024\/12\/dee42621cf58ffd.jpg\" alt=\"\" width=\"224\" height=\"420\" \/><\/p>\n<h3>3. KamradtSemanticChunker<\/h3>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"\" src=\"https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2024\/12\/a666c51992438a4.jpg\" alt=\"\" width=\"264\" height=\"477\" \/><\/p>\n<h3>4. KamradtModifiedChunker<\/h3>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"\" src=\"https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2024\/12\/9969b79a613b781.jpg\" alt=\"\" width=\"257\" height=\"677\" \/><\/p>\n<h3>5. ClusterSemanticChunker<\/h3>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"\" src=\"https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2024\/12\/e9e08faccb6ab7e.jpg\" alt=\"\" width=\"245\" height=\"707\" \/><\/p>\n<h3>6. LLMSemanticChunker<\/h3>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"\" src=\"https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2024\/12\/b16d43b5f4e9fd6.jpg\" alt=\"\" width=\"225\" height=\"614\" \/><\/p>\n<p>&nbsp;<\/p>\n<h2>\u65b9\u6cd5\u63cf\u8ff0<\/h2>\n<ol>\n<li><strong>RecursiveCharacterTextSplitter<\/strong>\uff1a\u57fa\u4e8e\u5206\u9694\u7b26\u7684\u5c42\u6b21\u7ed3\u6784\u5206\u5272\u6587\u672c\uff0c\u4f18\u5148\u8003\u8651\u6587\u6863\u4e2d\u7684\u81ea\u7136\u65ad\u70b9\u3002<\/li>\n<li><strong>TokenTextSplitter<\/strong>\uff1a\u5c06\u6587\u672c\u5206\u5272\u4e3a\u56fa\u5b9a\u6570\u91cf Token \u7684\u5757\uff0c\u786e\u4fdd\u5206\u5272\u53d1\u751f\u5728 Token \u8fb9\u754c\u3002<\/li>\n<li><strong>KamradtSemanticChunker<\/strong>\uff1a\u4f7f\u7528\u6ed1\u52a8\u7a97\u53e3\u5d4c\u5165\u8bc6\u522b\u8bed\u4e49\u4e0d\u8fde\u7eed\u70b9\uff0c\u5e76\u636e\u6b64\u5206\u5272\u6587\u672c\u3002<\/li>\n<li><strong>KamradtModifiedChunker<\/strong>\uff1a\u6539\u8fdb\u7248\u7684 KamradtSemanticChunker\uff0c\u4f7f\u7528\u4e8c\u5206\u641c\u7d22\u627e\u5230\u5206\u5272\u7684\u6700\u4f73\u9608\u503c\u3002<\/li>\n<li><strong>ClusterSemanticChunker<\/strong>\uff1a\u5c06\u6587\u672c\u5206\u5272\u6210\u5c0f\u5757\uff0c\u8ba1\u7b97\u5d4c\u5165\uff0c\u5e76\u4f7f\u7528\u52a8\u6001\u89c4\u5212\u6839\u636e\u8bed\u4e49\u76f8\u4f3c\u6027\u521b\u5efa\u6700\u4f18\u5206\u5757\u3002<\/li>\n<li><strong>LLMSemanticChunker<\/strong>\uff1a\u5229\u7528\u8bed\u8a00\u6a21\u578b\u786e\u5b9a\u6587\u672c\u4e2d\u9002\u5f53\u7684\u5206\u5272\u70b9\u3002<\/li>\n<\/ol>\n<h2>\u4f7f\u7528\u65b9\u6cd5<\/h2>\n<p>\u8981\u5728\u60a8\u7684 RAG \u6d41\u7a0b\u4e2d\u4f7f\u7528\u8fd9\u4e9b\u5206\u5757\u65b9\u6cd5\uff1a<\/p>\n<ol>\n<li>\u4ece\u00a0<code>chunkers<\/code>\u00a0\u6a21\u5757\u4e2d\u5bfc\u5165\u6240\u9700\u7684\u5206\u5757\u5668\u3002<\/li>\n<li>\u4f7f\u7528\u9002\u5f53\u7684\u53c2\u6570\u521d\u59cb\u5316\u5206\u5757\u5668\uff08\u4f8b\u5982\uff0c\u6700\u5927\u5206\u5757\u5927\u5c0f\u3001\u91cd\u53e0\uff09\u3002<\/li>\n<li>\u5c06\u60a8\u7684\u6587\u6863\u4f20\u9012\u7ed9\u5206\u5757\u5668\u4ee5\u83b7\u53d6\u5206\u5757\u7ed3\u679c\u3002<\/li>\n<\/ol>\n<p>\u793a\u4f8b\uff1a<\/p>\n<pre><code>from chunkers import RecursiveCharacterTextSplitter\r\nchunker = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=200)\r\nchunks = chunker.split_text(your_document)\r\n<\/code><\/pre>\n<h2>\u5982\u4f55\u9009\u62e9\u5206\u5757\u65b9\u6cd5<\/h2>\n<p>\u9009\u62e9\u5206\u5757\u65b9\u6cd5\u53d6\u51b3\u4e8e\u60a8\u7684\u5177\u4f53\u7528\u4f8b\uff1a<\/p>\n<ul>\n<li>\u5bf9\u4e8e\u7b80\u5355\u7684\u6587\u672c\u5206\u5272\uff0c\u53ef\u4ee5\u4f7f\u7528 RecursiveCharacterTextSplitter \u6216 TokenTextSplitter\u3002<\/li>\n<li>\u5982\u679c\u9700\u8981\u8bed\u4e49\u611f\u77e5\u7684\u5206\u5272\uff0c\u53ef\u4ee5\u8003\u8651 KamradtSemanticChunker \u6216 KamradtModifiedChunker\u3002<\/li>\n<li>\u5bf9\u4e8e\u66f4\u9ad8\u7ea7\u7684\u8bed\u4e49\u5206\u5757\uff0c\u8bf7\u4f7f\u7528 ClusterSemanticChunker \u6216 LLMSemanticChunker\u3002<\/li>\n<\/ul>\n<p>\u9009\u62e9\u65b9\u6cd5\u65f6\u9700\u8981\u8003\u8651\u7684\u56e0\u7d20\uff1a<\/p>\n<ul>\n<li>\u6587\u6863\u7ed3\u6784\u548c\u5185\u5bb9\u7c7b\u578b<\/li>\n<li>\u6240\u9700\u7684\u5206\u5757\u5927\u5c0f\u548c\u91cd\u53e0<\/li>\n<li>\u53ef\u7528\u7684\u8ba1\u7b97\u8d44\u6e90<\/li>\n<li>\u68c0\u7d22\u7cfb\u7edf\u7684\u7279\u5b9a\u8981\u6c42\uff08\u4f8b\u5982\uff0c\u57fa\u4e8e\u5411\u91cf\u8fd8\u662f\u57fa\u4e8e\u5173\u952e\u8bcd\uff09<\/li>\n<\/ul>\n<p>\u53ef\u4ee5\u5c1d\u8bd5\u4e0d\u540c\u7684\u65b9\u6cd5\uff0c\u627e\u5230\u6700\u9002\u5408\u60a8\u7684\u6587\u6863\u548c\u68c0\u7d22\u9700\u6c42\u7684\u65b9\u6cd5\u3002<\/p>\n<h2>\u4e0e RAG \u7cfb\u7edf\u7684\u96c6\u6210<\/h2>\n<p>\u5b8c\u6210\u5206\u5757\u540e\uff0c\u901a\u5e38\u4f1a\u6267\u884c\u4ee5\u4e0b\u6b65\u9aa4\uff1a<\/p>\n<ol>\n<li>\u4e3a\u6bcf\u4e2a\u5206\u5757\u751f\u6210\u5d4c\u5165\uff08\u9002\u7528\u4e8e\u57fa\u4e8e\u5411\u91cf\u7684\u68c0\u7d22\u7cfb\u7edf\uff09\u3002<\/li>\n<li>\u5728\u9009\u5b9a\u7684\u68c0\u7d22\u7cfb\u7edf\u4e2d\u7d22\u5f15\u8fd9\u4e9b\u5206\u5757\uff08\u4f8b\u5982\uff0c\u5411\u91cf\u6570\u636e\u5e93\u3001\u5012\u6392\u7d22\u5f15\uff09\u3002<\/li>\n<li>\u5728\u56de\u7b54\u67e5\u8be2\u65f6\uff0c\u4f7f\u7528\u68c0\u7d22\u6b65\u9aa4\u4e2d\u7684\u7d22\u5f15\u5206\u5757\u3002<\/li>\n<\/ol>\n<p>&nbsp;<\/p>\n<p><img decoding=\"async\" src=\"https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2024\/12\/2bba088276f3664-scaled.jpg\" alt=\"\" \/><\/p>\n","protected":false},"excerpt":{"rendered":"<p>\u7b80\u4ecb \u6570\u636e\u5206\u5757\u662f\u68c0\u7d22\u589e\u5f3a\u751f\u6210\uff08RAG\uff09\u7cfb\u7edf\u4e2d\u7684\u5173\u952e\u6b65\u9aa4\u3002\u5b83\u5c06\u5927\u578b\u6587\u6863\u5206\u89e3\u4e3a\u8f83\u5c0f\u7684\u3001\u53ef\u7ba1\u7406\u7684\u7247\u6bb5\uff0c\u4ee5\u4fbf\u9ad8\u6548\u7d22\u5f15\u3001\u68c0\u7d22\u548c\u5904\u7406\u3002\u672c README \u63d0\u4f9b\u4e86 RAG \u7ba1\u9053\u4e2d\u53ef\u7528\u7684\u5404\u79cd\u5206\u5757\u65b9\u6cd5\u6982\u8ff0\u3002 https:\/\/github.com\/adithya&#8230;<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[34],"tags":[],"class_list":["post-14910","post","type-post","status-publish","format-standard","hentry","category-knowledge"],"_links":{"self":[{"href":"https:\/\/www.kdjingpai.com\/ja\/wp-json\/wp\/v2\/posts\/14910","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.kdjingpai.com\/ja\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.kdjingpai.com\/ja\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.kdjingpai.com\/ja\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.kdjingpai.com\/ja\/wp-json\/wp\/v2\/comments?post=14910"}],"version-history":[{"count":0,"href":"https:\/\/www.kdjingpai.com\/ja\/wp-json\/wp\/v2\/posts\/14910\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.kdjingpai.com\/ja\/wp-json\/wp\/v2\/media?parent=14910"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.kdjingpai.com\/ja\/wp-json\/wp\/v2\/categories?post=14910"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.kdjingpai.com\/ja\/wp-json\/wp\/v2\/tags?post=14910"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}