{"id":53709,"date":"2025-02-16T12:40:21","date_gmt":"2025-02-16T04:40:21","guid":{"rendered":"https:\/\/fwq.ai\/blog\/53709\/"},"modified":"2025-02-16T12:40:21","modified_gmt":"2025-02-16T04:40:21","slug":"deepseek-r1%e9%a9%b1%e5%8a%a8%e7%9a%84pdf%e8%a7%a3%e6%9e%90","status":"publish","type":"post","link":"https:\/\/fwq.ai\/blog\/53709\/","title":{"rendered":"DeepSeek R1\u9a71\u52a8\u7684PDF\u89e3\u6790"},"content":{"rendered":"<p>PDF \u89e3\u6790\u662f\u8bb8\u591a\u5e94\u7528\u7a0b\u5e8f\u7684\u5173\u952e\u4efb\u52a1\uff0c\u5c24\u5176\u662f\u5728\u68c0\u7d22\u589e\u5f3a\u751f\u6210 (RAG) \u7cfb\u7edf\u4e2d\u3002\u6bcf\u4e2a\u4eba\u90fd\u60f3\u4ece PDF \u4e2d\u63d0\u53d6\u7ed3\u6784\u5316\u6570\u636e\uff0c\u4f46\u8fd9\u975e\u5e38\u56f0\u96be\u3002\u7a0b\u5e8f\u5316\u89e3\u6790\u901a\u5e38\u4f1a\u5931\u8d25\uff0c\u56e0\u4e3a\u5b83\u4f9d\u8d56\u4e8e PDF \u7684\u5b8c\u7f8e\u7ed3\u6784\u2014\u2014\u8fd9\u5728\u73b0\u5b9e\u4e16\u754c\u7684\u6587\u6863\u4e2d\u5f88\u5c11\u89c1\u3002\u975e\u63a8\u7406 LLM \u53ef\u4ee5\u63d0\u4f9b\u5e2e\u52a9\uff0c\u4f46\u5b83\u4eec\u5bf9\u4e8e\u50cf\u8fd9\u6837\u7684\u590d\u6742\u4efb\u52a1\u5e76\u4e0d\u603b\u662f\u53ef\u9760\u7684\u3002<\/p>\n<p>\u56e0\u6b64\u6211\u4eec\u9700\u8981\u63a8\u7406\u6a21\u578b\u3002\u8fd9\u4e9b\u9ad8\u7ea7 LLM \u627f\u8bfa\u63d0\u4f9b\u66f4\u597d\u7684\u6027\u80fd\uff0c\u4f46\u5b83\u4eec\u6709\u4e00\u4e2a\u7f3a\u70b9\uff1a\u5b83\u4eec\u5f88\u6602\u8d35\u3002\u4f8b\u5982\uff0cOpenAI \u7684 o1 \u6bcf 100 \u4e07\u4e2a token \u7684\u6210\u672c\u4e3a 60 \u7f8e\u5143\u3002<\/p>\n<p>\u53e6\u4e00\u65b9\u9762\uff0cDeepSeek \u7684\u65b0\u7248\u672c R1 \u63d0\u4f9b\u7684\u63a8\u7406\u6a21\u578b\u6bcf 100 \u4e07\u4e2a token \u7684\u6210\u672c\u4ec5\u4e3a 2 \u7f8e\u5143\u2014\u2014\u8fd9\u53ea\u662f\u6210\u672c\u7684\u4e00\u200b\u200b\u5c0f\u90e8\u5206\u3002\u81ea\u7136\uff0c\u6211\u60f3\u770b\u770b\u5b83\u4eec\u5728\u771f\u5b9e\u573a\u666f\u4e2d\u7684\u8868\u73b0\uff1a\u89e3\u6790\u4e00\u4efd\u6df7\u4e71\u3001\u975e\u7ed3\u6784\u5316\u7684\u6d4b\u8bd5\u62a5\u544a PDF\u3002<\/p>\n<h2>1\u3001\u6570\u636e\uff1a\u6d4b\u8bd5\u62a5\u544a<\/h2>\n<p>\u8f93\u5165\u6587\u4ef6\u662f\u4e00\u4efd\u6d4b\u8bd5\u62a5\u544a\u3002\u51fa\u4e8e\u4fdd\u5bc6\u6027\u8003\u8651\uff0c\u6211\u5ba1\u67e5\u4e86\u51e0\u4e4e\u6240\u6709\u5185\u5bb9\uff0c\u4f46\u4f60\u4ecd\u7136\u53ef\u4ee5\u7406\u89e3\u6df7\u4e71\u7684\u5e03\u5c40<\/p>\n<h2>2\u3001\u65b9\u6cd5\uff1a\u9884\u5904\u7406\u548c\u63a8\u7406 LLM<\/h2>\n<h3>2.1 \u4f7f\u7528\u7a7a\u95f4\u5206\u6790\u8fdb\u884c\u9884\u5904\u7406<\/h3>\n<p>\u7b2c\u4e00\u6b65\u662f\u4f7f\u7528 pdfplumber \u4ece PDF \u4e2d\u63d0\u53d6\u6587\u672c\u5757\u3002\u6bcf\u4e2a\u6587\u672c\u5757\u90fd\u5e26\u6709\u5176\u5750\u6807\uff0c\u8fd9\u4f7f\u6211\u80fd\u591f\u5206\u6790\u5b83\u4eec\u7684\u7a7a\u95f4\u5173\u7cfb\u3002<\/p>\n<pre><code>import pdfplumber\n\nwith pdfplumber.open(\"report.pdf\") as pdf:\n    for page in pdf.pages:\n        for block in page.extract_words():\n            print(block)<\/code><\/pre>\n<p>\u63a5\u4e0b\u6765\uff0c\u6211\u5728\u6bcf\u4e2a\u6587\u672c\u5757\u5468\u56f4\u521b\u5efa\u4e86\u4e00\u4e2a\u201c\u7a97\u53e3\u201d\u6765\u6355\u83b7\u5176\u5468\u56f4\u7684\u4e0a\u4e0b\u6587\u3002\u5bf9\u4e8e\u6bcf\u5bf9\u5757\uff0c\u6211\u8ba1\u7b97\u4e86\u5b83\u4eec\u4e4b\u95f4\u7684\u8ddd\u79bb\u548c\u89d2\u5ea6\u3002\u89d2\u5ea6\u6709\u52a9\u4e8e\u786e\u5b9a\u5757\u662f\u6c34\u5e73\u5bf9\u9f50\uff08\u53ef\u80fd\u662f\u540c\u4e00\u884c\u7684\u4e00\u90e8\u5206\uff09\u8fd8\u662f\u5782\u76f4\u5bf9\u9f50\uff08\u53ef\u80fd\u662f\u4e00\u5217\u7684\u4e00\u90e8\u5206\uff09\u3002<\/p>\n<pre><code>import math\n\ndef calculate_distance_and_angle(block1, block2):\n    x1, y1 = block1['x0'], block1['top']\n    x2, y2 = block2['x0'], block2['top']\n    dx, dy = x2 - x1, y2 - y1\n    distance = math.hypot(dx, dy)\n    angle = math.degrees(math.atan2(dy, dx)) % 360\n    return distance, angle<\/code><\/pre>\n<p>\u5229\u7528\u8fd9\u4e9b\u4fe1\u606f\uff0c\u6211\u5728\u5f7c\u6b64\u8db3\u591f\u63a5\u8fd1\u7684\u5757\u4e4b\u95f4\u521b\u5efa\u4e86\u552f\u4e00\u7684\u201c\u94fe\u63a5\u201d\u3002\u6bcf\u4e2a\u94fe\u63a5\u90fd\u5305\u62ec\u8d77\u59cb\u5757\u3001\u7ec8\u6b62\u5757\u3001\u8ddd\u79bb\u548c\u89d2\u5ea6\u3002\u8fd9\u4e9b\u94fe\u63a5\u5b58\u50a8\u5728 markdown \u8868\u4e2d\uff0c\u4ee5\u4fbf\u4e8e\u53c2\u8003\u3002<\/p>\n<table>\n<thead>\n<tr>\n<th>Start Block<\/th>\n<th>End Block<\/th>\n<th>Distance<\/th>\n<th>Angle<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>Project No.<\/td>\n<td>2025-001<\/td>\n<td>16<\/td>\n<td>0<\/td>\n<\/tr>\n<tr>\n<td>Client \/ Owner<\/td>\n<td>Global Company<\/td>\n<td>14<\/td>\n<td>10<\/td>\n<\/tr>\n<tr>\n<td>Name of Item<\/td>\n<td>Example Item<\/td>\n<td>15<\/td>\n<td>5<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<h3>2.2 \u7528\u63a8\u7406LLM\u83b7\u5f97\u7ed3\u6784\u5316\u8f93\u51fa<\/h3>\n<p>\u6709\u4e86\u9884\u5904\u7406\u7684\u6570\u636e\uff0c\u6211\u4e3a LLM \u7f16\u5199\u4e86\u4e00\u6761\u4eba\u5de5\u6d88\u606f\u3002\u6b64\u6d88\u606f\u5305\u62ec\uff1a<\/p>\n<ul>\n<li>\u63cf\u8ff0\u4efb\u52a1\u7684\u7cfb\u7edf\u63d0\u793a\u3002<\/li>\n<li>\u5b9a\u4e49\u6240\u9700\u8f93\u51fa\u683c\u5f0f\u7684 GraphQL \u6a21\u5f0f\u3002<\/li>\n<li>\u94fe\u63a5\u7684 markdown \u8868\u3002<\/li>\n<\/ul>\n<p>\u8fd9\u662f\u6211\u4f7f\u7528\u7684 GraphQL \u6a21\u5f0f\u7684\u793a\u4f8b\uff1a<\/p>\n<pre><code>type TestReport {\n  projectNo: String\n  reportNo: String\n  client: String\n  itemName: String\n  dateOfInspection: Date\n  location: String\n  ...\n}<\/code><\/pre>\n<p>\u6211\u5c06\u6b64\u6d88\u606f\u53d1\u9001\u7ed9\u4e86\u4e24\u4e2a LLM\uff1aDeepSeek-R1 \u548c OpenAI o1\u3002\u7ed3\u679c\u4ee4\u4eba\u60ca\u8bb6\u3002<\/p>\n<h2>3\u3001\u7ed3\u675f\u8bed<\/h2>\n<p>DeepSeek \u5b8c\u6210\u4e86\u4efb\u52a1\u3002\u5b83\u6b63\u786e\u5730\u6620\u5c04\u4e86\u6240\u6709\u5b57\u6bb5\uff0c\u5305\u62ec <code>inspectedBy<\/code>\u548c <code>reviewedBy<\/code>\u7b49\u68d8\u624b\u7684\u5b57\u6bb5\uff0c\u5e76\u4ea7\u751f\u4e86\u5e72\u51c0\u3001\u7ed3\u6784\u5316\u7684\u8f93\u51fa\u3002\u53e6\u4e00\u65b9\u9762\uff0cOpenAI GPT-4 \u6df7\u6dc6\u4e86\u4e00\u4e9b\u5b57\u6bb5\u5e76\u9057\u6f0f\u4e86\u4e00\u4e9b\u7ec6\u8282\uff0c\u5c3d\u7ba1\u4ef7\u683c\u9ad8\u51fa 30 \u500d\u3002<\/p>\n<p>\u6211\u4ece\u8fd9\u6b21\u7ecf\u5386\u4e2d\u5f97\u5230\u4e86\u4e24\u70b9\uff1a<\/p>\n<ul>\n<li>\u7a7a\u95f4\u5206\u6790\u529f\u80fd\u5f3a\u5927\u3002\u9ad8\u8d28\u91cf\u7684\u6570\u636e\u59cb\u7ec8\u81f3\u5173\u91cd\u8981\u3002\u901a\u8fc7\u9884\u5904\u7406\u6570\u636e\u548c\u8ba1\u7b97\u7a7a\u95f4\u5173\u7cfb\uff0c\u6211\u80fd\u591f\u4e3a LLM \u63d0\u4f9b\u6709\u610f\u4e49\u7684\u80cc\u666f\u4fe1\u606f\u3002<\/li>\n<li>DeepSeek \u4e0d\u662f\u5728\u80e1\u95f9\u3002R1 \u53ef\u80fd\u4f1a\u6539\u53d8\u8fd9\u7c7b\u4efb\u52a1\u7684\u6e38\u620f\u89c4\u5219\u3002\u5b83\u7684\u7ecf\u6d4e\u6027\u548c\u51c6\u786e\u6027\u4f7f\u5176\u6210\u4e3a\u4efb\u4f55\u5904\u7406\u975e\u7ed3\u6784\u5316\u6570\u636e\u7684\u4eba\u7684\u6709\u529b\u9009\u62e9\u3002<\/li>\n<\/ul>\n<p>\u5982\u679c\u4f60\u6b63\u5728\u52aa\u529b\u5e94\u5bf9 PDF \u89e3\u6790\u6216\u7c7b\u4f3c\u7684\u6311\u6218\uff0c\u90a3\u4e48\u503c\u5f97\u5c1d\u8bd5\u4e00\u4e0b DeepSeek\u3002\u4f60\u53ef\u80fd\u4f1a\u50cf\u6211\u4e00\u6837\u53d1\u73b0\uff0c\u6700\u597d\u7684\u89e3\u51b3\u65b9\u6848\u5e76\u4e0d\u603b\u662f\u6700\u6602\u8d35\u7684\u3002<\/p>\n<hr>\n","protected":false},"excerpt":{"rendered":"<p>PDF \u89e3\u6790\u662f\u8bb8\u591a\u5e94\u7528\u7a0b\u5e8f\u7684\u5173\u952e\u4efb\u52a1\uff0c\u5c24\u5176\u662f\u5728\u68c0\u7d22\u589e\u5f3a\u751f\u6210 (RAG) \u7cfb\u7edf\u4e2d\u3002\u6bcf\u4e2a\u4eba\u90fd\u60f3\u4ece PDF \u4e2d\u63d0\u53d6\u7ed3\u6784\u5316\u6570\u636e\uff0c\u4f46\u8fd9\u975e\u5e38\u56f0\u96be\u3002\u7a0b\u5e8f\u5316\u89e3\u6790\u901a\u5e38\u4f1a\u5931\u8d25\uff0c\u56e0\u4e3a\u5b83\u4f9d\u8d56\u4e8e PDF \u7684\u5b8c\u7f8e\u7ed3\u6784\u2014\u2014\u8fd9\u5728\u73b0\u5b9e\u4e16\u754c\u7684\u6587\u6863\u4e2d\u5f88\u5c11\u89c1\u3002\u975e\u63a8\u7406 LLM \u53ef\u4ee5\u63d0\u4f9b\u5e2e\u52a9\uff0c\u4f46\u5b83\u4eec\u5bf9\u4e8e\u50cf\u8fd9\u6837\u7684\u590d\u6742\u4efb\u52a1\u5e76\u4e0d\u603b\u662f\u53ef\u9760\u7684\u3002 \u56e0\u6b64\u6211\u4eec\u9700\u8981\u63a8\u7406\u6a21\u578b\u3002\u8fd9\u4e9b\u9ad8\u7ea7 LLM \u627f\u8bfa\u63d0\u4f9b\u66f4\u597d\u7684\u6027\u80fd\uff0c\u4f46\u5b83\u4eec\u6709\u4e00\u4e2a\u7f3a\u70b9\uff1a\u5b83\u4eec\u5f88\u6602\u8d35\u3002\u4f8b\u5982\uff0cOpenAI \u7684 o1 \u6bcf 100 \u4e07\u4e2a token \u7684\u6210\u672c\u4e3a 60 \u7f8e\u5143\u3002 \u53e6\u4e00\u65b9\u9762\uff0cDeepSeek \u7684\u65b0\u7248\u672c R1 \u63d0\u4f9b\u7684\u63a8\u7406\u6a21\u578b\u6bcf 100 \u4e07\u4e2a token \u7684\u6210\u672c\u4ec5\u4e3a 2 \u7f8e\u5143\u2014\u2014\u8fd9\u53ea\u662f\u6210\u672c\u7684\u4e00\u200b\u200b\u5c0f\u90e8\u5206\u3002\u81ea\u7136\uff0c\u6211\u60f3\u770b\u770b\u5b83\u4eec\u5728\u771f\u5b9e\u573a\u666f\u4e2d\u7684\u8868\u73b0\uff1a\u89e3\u6790\u4e00\u4efd\u6df7\u4e71\u3001\u975e\u7ed3\u6784\u5316\u7684\u6d4b\u8bd5\u62a5\u544a PDF\u3002 1\u3001\u6570\u636e\uff1a\u6d4b\u8bd5\u62a5\u544a \u8f93\u5165\u6587\u4ef6\u662f\u4e00\u4efd\u6d4b\u8bd5\u62a5\u544a\u3002\u51fa\u4e8e\u4fdd\u5bc6\u6027\u8003\u8651\uff0c\u6211\u5ba1\u67e5\u4e86\u51e0\u4e4e\u6240\u6709\u5185\u5bb9\uff0c\u4f46\u4f60\u4ecd\u7136\u53ef\u4ee5\u7406\u89e3\u6df7\u4e71\u7684\u5e03\u5c40 2\u3001\u65b9\u6cd5\uff1a\u9884\u5904\u7406\u548c\u63a8\u7406 LLM 2.1 \u4f7f\u7528\u7a7a\u95f4\u5206\u6790\u8fdb\u884c\u9884\u5904\u7406 \u7b2c\u4e00\u6b65\u662f\u4f7f\u7528 pdfplumber \u4ece PDF \u4e2d\u63d0\u53d6\u6587\u672c\u5757\u3002\u6bcf\u4e2a\u6587\u672c\u5757\u90fd\u5e26\u6709\u5176\u5750\u6807\uff0c\u8fd9\u4f7f\u6211\u80fd\u591f\u5206\u6790\u5b83\u4eec\u7684\u7a7a\u95f4\u5173\u7cfb\u3002 import pdfplumber with pdfplumber.open(&#8220;report.pdf&#8221;) as pdf: for page in pdf.pages: for [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[13],"tags":[],"class_list":["post-53709","post","type-post","status-publish","format-standard","hentry","category-ai"],"_links":{"self":[{"href":"https:\/\/fwq.ai\/blog\/wp-json\/wp\/v2\/posts\/53709","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/fwq.ai\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/fwq.ai\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/fwq.ai\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/fwq.ai\/blog\/wp-json\/wp\/v2\/comments?post=53709"}],"version-history":[{"count":0,"href":"https:\/\/fwq.ai\/blog\/wp-json\/wp\/v2\/posts\/53709\/revisions"}],"wp:attachment":[{"href":"https:\/\/fwq.ai\/blog\/wp-json\/wp\/v2\/media?parent=53709"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/fwq.ai\/blog\/wp-json\/wp\/v2\/categories?post=53709"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/fwq.ai\/blog\/wp-json\/wp\/v2\/tags?post=53709"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}