{"id":1320,"date":"2023-02-25T17:22:28","date_gmt":"2023-02-25T09:22:28","guid":{"rendered":"https:\/\/www.aqwu.net\/wp\/?p=1320"},"modified":"2023-02-25T19:22:32","modified_gmt":"2023-02-25T11:22:32","slug":"openai-api-%e6%8c%87%e5%8d%97%ef%bc%9a5-%e5%b5%8c%e5%85%a5","status":"publish","type":"post","link":"https:\/\/www.aqwu.net\/wp\/?p=1320","title":{"rendered":"OpenAI API \u6307\u5357\uff1a5.\u5d4c\u5165"},"content":{"rendered":"\n<p><a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\/what-are-embeddings\"><\/a><\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\/what-are-embeddings\">\u4ec0\u4e48\u662f\u5d4c\u5165\uff1f<\/a><\/h2>\n\n\n\n<p>OpenAI\u7684\u6587\u672c\u5d4c\u5165\u8861\u91cf\u6587\u672c\u5b57\u7b26\u4e32\u7684\u76f8\u5173\u6027\u3002\u5d4c\u5165\u901a\u5e38\u7528\u4e8e\uff1a<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>\u641c\u7d22<\/strong>\uff08\u5176\u4e2d\u7ed3\u679c\u6309\u4e0e\u67e5\u8be2\u5b57\u7b26\u4e32\u7684\u76f8\u5173\u6027\u6392\u540d\uff09<\/li>\n\n\n\n<li><strong>\u805a\u7c7b\u5206\u6790<\/strong>\uff08\u5176\u4e2d\u6587\u672c\u5b57\u7b26\u4e32\u6309\u76f8\u4f3c\u6027\u5206\u7ec4\uff09<\/li>\n\n\n\n<li>\u5efa\u8bae<strong>\uff08\u5efa\u8bae<\/strong>\u4f7f\u7528\u5177\u6709\u76f8\u5173\u6587\u672c\u5b57\u7b26\u4e32\u7684\u9879\u76ee\uff09<\/li>\n\n\n\n<li><strong>\u5f02\u5e38\u68c0\u6d4b<\/strong>\uff08\u8bc6\u522b\u76f8\u5173\u6027\u4e0d\u5927\u7684\u5f02\u5e38\u503c\uff09<\/li>\n\n\n\n<li><strong>\u591a\u6837\u6027\u6d4b\u91cf<\/strong>\uff08\u5206\u6790\u76f8\u4f3c\u6027\u5206\u5e03\uff09<\/li>\n\n\n\n<li><strong>\u5206\u7c7b<\/strong>\uff08\u5176\u4e2d\u6587\u672c\u5b57\u7b26\u4e32\u6309\u5176\u6700\u76f8\u4f3c\u7684\u6807\u7b7e\u5206\u7c7b\uff09<\/li>\n<\/ul>\n\n\n\n<p>\u539f\u6587\u94fe\u63a5\uff1a<a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\">Embeddings &#8211; OpenAI API<\/a><\/p>\n\n\n\n<p>\u5d4c\u5165\u662f\u6d6e\u70b9\u6570\u7684\u5411\u91cf\uff08\u5217\u8868\uff09\u3002\u4e24\u4e2a\u5411\u91cf\u4e4b\u95f4\u7684<a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\/which-distance-function-should-i-use\">\u8ddd\u79bb<\/a>\u8861\u91cf\u5b83\u4eec\u7684\u76f8\u5173\u6027\u3002\u5c0f\u8ddd\u79bb\u8868\u793a\u9ad8\u76f8\u5173\u6027\uff0c\u5927\u8ddd\u79bb\u8868\u793a\u4f4e\u76f8\u5173\u6027\u3002<\/p>\n\n\n\n<p>\u8bbf\u95ee\u6211\u4eec\u7684<a href=\"https:\/\/openai.com\/api\/pricing\/\" target=\"_blank\" rel=\"noreferrer noopener\">\u5b9a\u4ef7\u9875\u9762<\/a>\uff0c\u4e86\u89e3\u5d4c\u5165\u5b9a\u4ef7\u3002\u8bf7\u6c42\u6839\u636e\u53d1\u9001<a href=\"https:\/\/platform.openai.com\/docs\/api-reference\/embeddings\/create#embeddings\/create-input\">\u7684\u8f93\u5165<\/a>\u4e2d\u7684<a href=\"https:\/\/platform.openai.com\/tokenizer\">\u4ee4\u724c<\/a>\u6570\u8ba1\u8d39\u3002<\/p>\n\n\n\n<p><strong>\u82e5\u8981\u67e5\u770b\u5d4c\u5165\u7684\u5b9e\u9645\u5e94\u7528\uff0c\u8bf7\u67e5\u770b\u6211\u4eec\u7684\u4ee3\u7801\u793a\u4f8b<\/strong><\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>\u5206\u7c7b<\/li>\n\n\n\n<li>\u4e3b\u9898\u805a\u7c7b<\/li>\n\n\n\n<li>\u641c\u7d22<\/li>\n\n\n\n<li>\u5efa\u8bae<\/li>\n<\/ul>\n\n\n\n<p><a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\/use-cases\">\u6d4f\u89c8\u793a\u4f8b<\/a><\/p>\n\n\n\n<p><a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\/how-to-get-embeddings\"><\/a><\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\/how-to-get-embeddings\">\u5982\u4f55\u83b7\u53d6\u5d4c\u5165<\/a><\/h2>\n\n\n\n<p>\u8981\u83b7\u53d6\u5d4c\u5165\uff0c\u8bf7\u5c06\u6587\u672c\u5b57\u7b26\u4e32\u53d1\u9001\u5230<a href=\"https:\/\/platform.openai.com\/docs\/api-reference\/embeddings\">\u5d4c\u5165 API \u7ec8\u7ed3\u70b9<\/a>\uff0c\u540c\u65f6\u9009\u62e9\u5d4c\u5165\u6a21\u578b ID\uff08\u4f8b\u5982\uff0c\uff09\u3002\u54cd\u5e94\u5c06\u5305\u542b\u4e00\u4e2a\u5d4c\u5165\uff0c\u60a8\u53ef\u4ee5\u63d0\u53d6\u3001\u4fdd\u5b58\u548c\u4f7f\u7528\u5d4c\u5165\u3002<code>text-embedding-ada-002<\/code><\/p>\n\n\n\n<p>\u793a\u4f8b\u8bf7\u6c42\uff1a<\/p>\n\n\n\n<p>\u793a\u4f8b\uff1a\u83b7\u53d6\u5d4c\u5165<\/p>\n\n\n\n<p>\u5377\u66f2\u9009\u62e9\u5e93\u87d2\u5377\u66f2\u590d\u5236<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>curl https:\/\/api.openai.com\/v1\/embeddings \\\n  -H \"Content-Type: application\/json\" \\\n  -H \"Authorization: Bearer $OPENAI_API_KEY\" \\\n  -d '{\"input\": \"Your text string goes here\",\n       \"model\":\"text-embedding-ada-002\"}'<\/code><\/pre>\n\n\n\n<p>\u793a\u4f8b\u54cd\u5e94\uff1a<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>{\n  \"data\": &#91;\n    {\n      \"embedding\": &#91;\n        -0.006929283495992422,\n        -0.005336422007530928,\n        ...\n        -4.547132266452536e-05,\n        -0.024047505110502243\n      ],\n      \"index\": 0,\n      \"object\": \"embedding\"\n    }\n  ],\n  \"model\": \"text-embedding-ada-002\",\n  \"object\": \"list\",\n  \"usage\": {\n    \"prompt_tokens\": 5,\n    \"total_tokens\": 5\n  }\n}<\/code><\/pre>\n\n\n\n<p>\u5728&nbsp;<a href=\"https:\/\/github.com\/openai\/openai-cookbook\/\" target=\"_blank\" rel=\"noreferrer noopener\">OpenAI Cookbook<\/a>&nbsp;\u4e2d\u67e5\u770b\u66f4\u591a Python \u4ee3\u7801\u793a\u4f8b\u3002<\/p>\n\n\n\n<p>\u4f7f\u7528 OpenAI \u5d4c\u5165\u65f6\uff0c\u8bf7\u8bb0\u4f4f\u5b83\u4eec\u7684<a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\/limitations-risks\">\u5c40\u9650\u6027\u548c\u98ce\u9669<\/a>\u3002<\/p>\n\n\n\n<p><a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\/embedding-models\"><\/a><\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\/embedding-models\">\u5d4c\u5165\u6a21\u578b<\/a><\/h2>\n\n\n\n<p>OpenAI \u63d0\u4f9b 16 \u4e2a\u7b2c\u4e8c\u4ee3\u5d4c\u5165\u6a21\u578b\uff08\u5728\u6a21\u578b ID \u4e2d\u8868\u793a<code>-002<\/code>\uff09\u548c 16 \u7b2c\u4e00\u4ee3\u6a21\u578b\uff08\u5728\u6a21\u578b ID \u4e2d\u8868\u793a<code>-001<\/code>\uff09\u3002<\/p>\n\n\n\n<p>\u6211\u4eec\u5efa\u8bae\u5bf9\u51e0\u4e4e\u6240\u6709\u7528\u4f8b\u4f7f\u7528\u6587\u672c\u5d4c\u5165-ada-002\u3002\u5b83\u66f4\u597d\u3001\u66f4\u4fbf\u5b9c\u3001\u66f4\u6613\u4e8e\u4f7f\u7528\u3002\u9605\u8bfb<a href=\"https:\/\/openai.com\/blog\/new-and-improved-embedding-model\" target=\"_blank\" rel=\"noreferrer noopener\">\u535a\u5ba2\u6587\u7ae0\u516c\u544a<\/a>\u3002<\/p>\n\n\n\n<figure class=\"wp-block-table\"><table><thead><tr><th>\u6a21\u578b\u751f\u6210<\/th><th>\u5206\u8bcd\u5668<\/th><th>\u6700\u5927\u8f93\u5165\u4ee4\u724c\u6570<\/th><th>\u77e5\u8bc6\u622a\u6b62<\/th><\/tr><\/thead><tbody><tr><td>V2 \u7248<\/td><td>cl100k_base<\/td><td>8191<\/td><td>Sep 2021<\/td><\/tr><tr><td>V1 \u7248<\/td><td>GPT-2\/GPT-3<\/td><td>2046<\/td><td>Aug 2020<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n<p>\u4f7f\u7528\u91cf\u6309\u6bcf\u4e2a\u8f93\u5165\u4ee4\u724c\u5b9a\u4ef7\uff0c\u8d39\u7387\u4e3a\u6bcf 0 \u4e2a\u4ee4\u724c 0004.1000 USD\uff0c\u6216\u6bcf\u7f8e\u5143\u7ea6 ~3\uff0c000 \u9875\uff08\u5047\u8bbe\u6bcf\u9875 ~800 \u4e2a\u4ee4\u724c\uff09\uff1a<\/p>\n\n\n\n<figure class=\"wp-block-table\"><table><thead><tr><th>\u578b<\/th><th>\u6bcf\u7f8e\u5143\u7c97\u7565\u9875\u6570<\/th><th><a href=\"https:\/\/paperswithcode.com\/sota\/zero-shot-text-search-on-beir\" target=\"_blank\" rel=\"noreferrer noopener\">BEIR<\/a>&nbsp;\u641c\u7d22\u8bc4\u4f30\u4e2d\u7684\u793a\u4f8b\u6027\u80fd<\/th><\/tr><\/thead><tbody><tr><td>text-embedding-ada-002<\/td><td>3000<\/td><td>53.9<\/td><\/tr><tr><td>*-davinci-*-001<\/td><td>6<\/td><td>52.8<\/td><\/tr><tr><td>*-curie-*-001<\/td><td>60<\/td><td>50.9<\/td><\/tr><tr><td>*-babbage-*-001<\/td><td>240<\/td><td>50.4<\/td><\/tr><tr><td>*-ada-*-001<\/td><td>300<\/td><td>49.0<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n<p><a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\/second-generation-models\"><\/a><\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\/second-generation-models\">\u7b2c\u4e8c\u4ee3\u8f66\u578b<\/a><\/h3>\n\n\n\n<figure class=\"wp-block-table\"><table><thead><tr><th>\u578b\u53f7\u540d\u79f0<\/th><th>\u5206\u8bcd\u5668<\/th><th>\u6700\u5927\u8f93\u5165\u4ee4\u724c\u6570<\/th><th>\u8f93\u51fa\u5c3a\u5bf8<\/th><\/tr><\/thead><tbody><tr><td>text-embedding-ada-002<\/td><td>cl100k_base<\/td><td>8191<\/td><td>1536<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n<p>\u7b2c\u4e00\u4ee3\u578b\u53f7\uff08\u4e0d\u63a8\u8350\uff09<\/p>\n\n\n\n<p><a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\/use-cases\"><\/a><\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\/use-cases\">\u4f7f\u7528\u6848\u4f8b<\/a><\/h2>\n\n\n\n<p>\u5728\u8fd9\u91cc\uff0c\u6211\u4eec\u5c55\u793a\u4e86\u4e00\u4e9b\u5177\u6709\u4ee3\u8868\u6027\u7684\u7528\u4f8b\u3002\u5bf9\u4e8e\u4ee5\u4e0b\u793a\u4f8b\uff0c\u6211\u4eec\u5c06\u4f7f\u7528<a href=\"https:\/\/www.kaggle.com\/snap\/amazon-fine-food-reviews\" target=\"_blank\" rel=\"noreferrer noopener\">\u4e9a\u9a6c\u900a\u7f8e\u98df\u8bc4\u8bba\u6570\u636e\u96c6<\/a>\u3002<\/p>\n\n\n\n<p><a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\/obtaining-the-embeddings\"><\/a><\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\/obtaining-the-embeddings\">\u83b7\u53d6\u5d4c\u5165<\/a><\/h3>\n\n\n\n<p>\u8be5\u6570\u636e\u96c6\u5305\u542b\u622a\u81f3 568 \u5e74 454 \u6708\u4e9a\u9a6c\u900a\u7528\u6237\u7559\u4e0b\u7684 2012\uff0c1 \u6761\u98df\u54c1\u8bc4\u8bba\u3002\u6211\u4eec\u5c06\u4f7f\u7528 000\uff0c&lt;&gt; \u6761\u6700\u65b0\u8bc4\u8bba\u7684\u5b50\u96c6\u8fdb\u884c\u8bf4\u660e\u3002\u8bc4\u8bba\u662f\u82f1\u6587\u7684\uff0c\u5f80\u5f80\u662f\u6b63\u9762\u6216\u8d1f\u9762\u7684\u3002\u6bcf\u6761\u8bc4\u8bba\u90fd\u6709\u4e00\u4e2a\u4ea7\u54c1 ID\u3001\u7528\u6237 ID\u3001\u5206\u6570\u3001\u8bc4\u8bba\u6807\u9898\uff08\u6458\u8981\uff09\u548c\u8bc4\u8bba\u6b63\u6587\uff08\u6587\u672c\uff09\u3002\u4f8b\u5982\uff1a<\/p>\n\n\n\n<figure class=\"wp-block-table\"><table><thead><tr><th>\u4ea7\u54c1\u7f16\u53f7<\/th><th>\u7528\u6237\u6807\u8bc6<\/th><th>\u5f97\u5206<\/th><th>\u603b\u7ed3<\/th><th>\u53d1\u77ed\u4fe1<\/th><\/tr><\/thead><tbody><tr><td>B001E4KFG0<\/td><td>A3SGXH7AUHU8GW<\/td><td>5<\/td><td>\u4f18\u8d28\u72d7\u7cae<\/td><td>\u6211\u4e70\u4e86\u51e0\u4e2a\u6d3b\u529b\u7f50\u5934&#8230;<\/td><\/tr><tr><td>B00813GRG4<\/td><td>A1D87F6ZCVE5NK<\/td><td>1<\/td><td>\u4e0d\u50cf\u5e7f\u544a\u4e0a\u6240\u8bf4\u7684\u90a3\u6837<\/td><td>\u4ea7\u54c1\u5230\u8fbe\u65f6\u6807\u6709\u5de8\u578b\u76d0\u6e0d\u82b1\u751f&#8230;<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n<p>\u6211\u4eec\u5c06\u8bc4\u8bba\u6458\u8981\u548c\u8bc4\u8bba\u6587\u672c\u5408\u5e76\u4e3a\u4e00\u4e2a\u7ec4\u5408\u6587\u672c\u3002\u8be5\u6a21\u578b\u5c06\u5bf9\u6b64\u7ec4\u5408\u6587\u672c\u8fdb\u884c\u7f16\u7801\u5e76\u8f93\u51fa\u5355\u4e2a\u5411\u91cf\u5d4c\u5165\u3002<a href=\"https:\/\/github.com\/openai\/openai-cookbook\/blob\/main\/examples\/Obtain_dataset.ipynb\" target=\"_blank\" rel=\"noreferrer noopener\">Obtain_dataset.ipynb<\/a><\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>def get_embedding(text, model=\"text-embedding-ada-002\"):\n   text = text.replace(\"\\n\", \" \")\n   return openai.Embedding.create(input = &#91;text], model=model)&#91;'data']&#91;0]&#91;'embedding']\n \ndf&#91;'ada_embedding'] = df.combined.apply(lambda x: get_embedding(x, model='text-embedding-ada-002'))\ndf.to_csv('output\/embedded_1k_reviews.csv', index=False)<\/code><\/pre>\n\n\n\n<p>\u8981\u4ece\u4fdd\u5b58\u7684\u6587\u4ef6\u52a0\u8f7d\u6570\u636e\uff0c\u53ef\u4ee5\u8fd0\u884c\u4ee5\u4e0b\u547d\u4ee4\uff1a<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>import pandas as pd\n \ndf = pd.read_csv('output\/embedded_1k_reviews.csv')\ndf&#91;'ada_embedding'] = df.ada_embedding.apply(eval).apply(np.array)<\/code><\/pre>\n\n\n\n<p><\/p>\n\n\n\n<p><a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\/limitations-risks\"><\/a><\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\/limitations-risks\">\u9650\u5236\u4e0e\u98ce\u9669<\/a><\/h2>\n\n\n\n<p>\u5728\u67d0\u4e9b\u60c5\u51b5\u4e0b\uff0c\u6211\u4eec\u7684\u5d4c\u5165\u6a21\u578b\u53ef\u80fd\u4e0d\u53ef\u9760\u6216\u6784\u6210\u793e\u4f1a\u98ce\u9669\uff0c\u5e76\u4e14\u5728\u6ca1\u6709\u7f13\u89e3\u63aa\u65bd\u7684\u60c5\u51b5\u4e0b\u53ef\u80fd\u4f1a\u9020\u6210\u4f24\u5bb3\u3002<\/p>\n\n\n\n<p><a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\/social-bias\"><\/a><\/p>\n\n\n\n<h4 class=\"wp-block-heading\"><a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\/social-bias\">\u793e\u4f1a\u504f\u89c1<\/a><\/h4>\n\n\n\n<blockquote class=\"wp-block-quote is-layout-flow wp-block-quote-is-layout-flow\">\n<p><strong>\u5c40\u9650\u6027<\/strong>\uff1a\u8fd9\u4e9b\u6a21\u578b\u7f16\u7801\u793e\u4f1a\u504f\u89c1\uff0c\u4f8b\u5982\u901a\u8fc7\u5bf9\u67d0\u4e9b\u7fa4\u4f53\u7684\u523b\u677f\u5370\u8c61\u6216\u8d1f\u9762\u60c5\u7eea\u3002<\/p>\n<\/blockquote>\n\n\n\n<p>\u6211\u4eec\u901a\u8fc7\u8fd0\u884cSEAT\uff08May\u7b49\u4eba\uff0c<a href=\"https:\/\/arxiv.org\/abs\/1903.10561\" target=\"_blank\" rel=\"noreferrer noopener\">2019<\/a>\uff09\u548cWinogender\uff08<a href=\"https:\/\/arxiv.org\/abs\/1804.09301\" target=\"_blank\" rel=\"noreferrer noopener\">Rudinger\u7b49\u4eba\uff0c2018<\/a>\uff09\u57fa\u51c6\u6d4b\u8bd5\u5728\u6211\u4eec\u7684\u6a21\u578b\u4e2d\u53d1\u73b0\u4e86\u504f\u89c1\u7684\u8bc1\u636e\u3002\u8fd9\u4e9b\u57fa\u51c6\u5171\u540c\u7531 7 \u4e2a\u6d4b\u8bd5\u7ec4\u6210\uff0c\u7528\u4e8e\u8861\u91cf\u6a21\u578b\u5728\u5e94\u7528\u4e8e\u6027\u522b\u540d\u79f0\u3001\u533a\u57df\u540d\u79f0\u548c\u4e00\u4e9b\u523b\u677f\u5370\u8c61\u65f6\u662f\u5426\u5305\u542b\u9690\u6027\u504f\u89c1\u3002<\/p>\n\n\n\n<p>\u4f8b\u5982\uff0c\u6211\u4eec\u53d1\u73b0\uff0c\u4e0e\u975e\u88d4\u7f8e\u56fd\u4eba\u7684\u540d\u5b57\u76f8\u6bd4\uff0c\u6211\u4eec\u7684\u6a21\u578b\u66f4\u5f3a\u70c8\u5730\u5c06\uff08a\uff09\u6b27\u6d32\u88d4\u7f8e\u56fd\u4eba\u7684\u540d\u5b57\u4e0e\u79ef\u6781\u7684\u60c5\u7eea\u8054\u7cfb\u8d77\u6765\uff0c\uff08b\uff09\u6d88\u6781\u7684\u523b\u677f\u5370\u8c61\u4e0e\u9ed1\u4eba\u5973\u6027\u8054\u7cfb\u8d77\u6765\u3002<\/p>\n\n\n\n<p>\u8fd9\u4e9b\u57fa\u51c6\u5728\u51e0\u4e2a\u65b9\u9762\u53d7\u5230\u9650\u5236\uff1a\uff08a\uff09\u5b83\u4eec\u53ef\u80fd\u65e0\u6cd5\u63a8\u5e7f\u5230\u60a8\u7684\u7279\u5b9a\u7528\u4f8b\uff0c\u4ee5\u53ca\uff08b\uff09\u5b83\u4eec\u4ec5\u6d4b\u8bd5\u4e00\u5c0f\u90e8\u5206\u53ef\u80fd\u7684\u793e\u4f1a\u504f\u89c1\u3002<\/p>\n\n\n\n<p><strong>\u8fd9\u4e9b\u6d4b\u8bd5\u662f\u521d\u6b65\u7684\uff0c\u6211\u4eec\u5efa\u8bae\u9488\u5bf9\u60a8\u7684\u7279\u5b9a\u7528\u4f8b\u8fd0\u884c\u6d4b\u8bd5\u3002<\/strong>\u8fd9\u4e9b\u7ed3\u679c\u5e94\u88ab\u89c6\u4e3a\u8be5\u73b0\u8c61\u5b58\u5728\u7684\u8bc1\u636e\uff0c\u800c\u4e0d\u662f\u60a8\u7684\u7528\u4f8b\u7684\u660e\u786e\u7279\u5f81\u3002\u8bf7\u53c2\u9605\u6211\u4eec\u7684<a href=\"https:\/\/platform.openai.com\/docs\/usage-policies\">\u4f7f\u7528\u653f\u7b56<\/a>\u4ee5\u83b7\u53d6\u66f4\u591a\u8be6\u7ec6\u4fe1\u606f\u548c\u6307\u5bfc\u3002<\/p>\n\n\n\n<p>\u5982\u679c\u60a8\u6709\u4efb\u4f55\u7591\u95ee\uff0c\u8bf7<a href=\"https:\/\/help.openai.com\/en\/\" target=\"_blank\" rel=\"noreferrer noopener\">\u901a\u8fc7\u804a\u5929\u8054\u7cfb\u6211\u4eec\u7684\u652f\u6301\u56e2\u961f<\/a>;\u6211\u4eec\u5f88\u4e50\u610f\u5c31\u6b64\u63d0\u4f9b\u5efa\u8bae\u3002<\/p>\n\n\n\n<p><a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\/blindness-to-recent-events\"><\/a><\/p>\n\n\n\n<h4 class=\"wp-block-heading\"><a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\/blindness-to-recent-events\">\u5bf9\u6700\u8fd1\u53d1\u751f\u7684\u4e8b\u4ef6\u89c6\u800c\u4e0d\u89c1<\/a><\/h4>\n\n\n\n<blockquote class=\"wp-block-quote is-layout-flow wp-block-quote-is-layout-flow\">\n<p><strong>\u9650\u5236<\/strong>\uff1a\u6a21\u578b\u7f3a\u4e4f\u5bf9 2020 \u5e74 &lt;&gt; \u6708\u4e4b\u540e\u53d1\u751f\u7684\u4e8b\u4ef6\u7684\u4e86\u89e3\u3002<\/p>\n<\/blockquote>\n\n\n\n<p>\u6211\u4eec\u7684\u6a21\u578b\u5728\u5305\u542b\u622a\u81f3 8 \u5e74 2020 \u6708\u7684\u771f\u5b9e\u4e16\u754c\u4e8b\u4ef6\u7684\u4e00\u4e9b\u4fe1\u606f\u7684\u6570\u636e\u96c6\u4e0a\u8fdb\u884c\u8bad\u7ec3\u3002\u5982\u679c\u60a8\u4f9d\u8d56\u4ee3\u8868\u6700\u8fd1\u4e8b\u4ef6\u7684\u6a21\u578b\uff0c\u90a3\u4e48\u5b83\u4eec\u53ef\u80fd\u65e0\u6cd5\u5f88\u597d\u5730\u6267\u884c\u3002<\/p>\n\n\n\n<p><a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\/frequently-asked-questions\"><\/a><\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\/frequently-asked-questions\">\u5e38\u89c1\u95ee\u9898<\/a><\/h2>\n\n\n\n<p><a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\/how-can-i-tell-how-many-tokens-a-string-has-before-i-embed-it\"><\/a><\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\/how-can-i-tell-how-many-tokens-a-string-has-before-i-embed-it\">\u5728\u5d4c\u5165\u5b57\u7b26\u4e32\u4e4b\u524d\uff0c\u5982\u4f55\u5224\u65ad\u5b57\u7b26\u4e32\u6709\u591a\u5c11\u4e2a\u6807\u8bb0\uff1f<\/a><\/h3>\n\n\n\n<p>\u5728Python\u4e2d\uff0c\u4f60\u53ef\u4ee5\u4f7f\u7528OpenAI\u7684tokenizer&nbsp;<a href=\"https:\/\/github.com\/openai\/tiktoken\" target=\"_blank\" rel=\"noreferrer noopener\"><code>tiktoken<\/code><\/a>\u5c06\u5b57\u7b26\u4e32\u62c6\u5206\u4e3a\u4ee3\u5e01\u3002<\/p>\n\n\n\n<p>\u793a\u4f8b\u4ee3\u7801\uff1a<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>import tiktoken\n\ndef num_tokens_from_string(string: str, encoding_name: str) -&gt; int:\n    \"\"\"Returns the number of tokens in a text string.\"\"\"\n    encoding = tiktoken.get_encoding(encoding_name)\n    num_tokens = len(encoding.encode(string))\n    return num_tokens\n\nnum_tokens_from_string(\"tiktoken is great!\", \"cl100k_base\")<\/code><\/pre>\n\n\n\n<p>\u5bf9\u4e8e\u7b2c\u4e8c\u4ee3\u5d4c\u5165\u6a21\u578b\uff0c\u5982 \uff0c\u8bf7\u4f7f\u7528\u7f16\u7801\u3002<code>text-embedding-ada-002<\/code><code>cl100k_base<\/code><\/p>\n\n\n\n<p>\u66f4\u591a\u8be6\u7ec6\u4fe1\u606f\u548c\u793a\u4f8b\u4ee3\u7801\u5728 OpenAI \u98df\u8c31\u6307\u5357\u5982\u4f55\u4f7f\u7528&nbsp;<a href=\"https:\/\/github.com\/openai\/openai-cookbook\/blob\/main\/examples\/How_to_count_tokens_with_tiktoken.ipynb\" target=\"_blank\" rel=\"noreferrer noopener\">tiktoken \u8ba1\u7b97\u4ee3\u5e01<\/a>\u3002<\/p>\n\n\n\n<p><a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\/how-can-i-retrieve-k-nearest-embedding-vectors-quickly\"><\/a><\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\/how-can-i-retrieve-k-nearest-embedding-vectors-quickly\">\u5982\u4f55\u5feb\u901f\u68c0\u7d22 K \u6700\u8fd1\u7684\u5d4c\u5165\u5411\u91cf\uff1f<\/a><\/h3>\n\n\n\n<p>\u4e3a\u4e86\u5feb\u901f\u641c\u7d22\u591a\u4e2a\u8f7d\u4f53\uff0c\u6211\u4eec\u5efa\u8bae\u4f7f\u7528\u77e2\u91cf\u6570\u636e\u5e93\u3002\u60a8\u53ef\u4ee5\u5728 GitHub \u4e0a\u7684<a href=\"https:\/\/github.com\/openai\/openai-cookbook\/tree\/main\/examples\/vector_databases\" target=\"_blank\" rel=\"noreferrer noopener\">\u98df\u8c31\u4e2d\u627e\u5230<\/a>\u4f7f\u7528\u77e2\u91cf\u6570\u636e\u5e93\u548c OpenAI API \u7684\u793a\u4f8b\u3002<\/p>\n\n\n\n<p>\u77e2\u91cf\u6570\u636e\u5e93\u9009\u9879\u5305\u62ec\uff1a<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><a rel=\"noreferrer noopener\" href=\"https:\/\/www.pinecone.io\/\" target=\"_blank\"><a rel=\"noreferrer noopener\" href=\"https:\/\/www.pinecone.io\/\" target=\"_blank\">Pinecone<\/a><\/a>\uff0c\u4e00\u4e2a\u5b8c\u5168\u6258\u7ba1\u7684\u8f7d\u4f53\u6570\u636e\u5e93<\/li>\n\n\n\n<li><a href=\"https:\/\/weaviate.io\/\" target=\"_blank\" rel=\"noreferrer noopener\">Weaviate<\/a>\uff0c\u4e00\u4e2a\u5f00\u6e90\u7684\u77e2\u91cf\u641c\u7d22\u5f15\u64ce<\/li>\n\n\n\n<li><a href=\"https:\/\/engineering.fb.com\/2017\/03\/29\/data-infrastructure\/faiss-a-library-for-efficient-similarity-search\/\" target=\"_blank\" rel=\"noreferrer noopener\">Faiss<\/a>\uff0cFacebook\u7684\u77e2\u91cf\u641c\u7d22\u7b97\u6cd5<\/li>\n\n\n\n<li><a rel=\"noreferrer noopener\" href=\"https:\/\/redis.com\/blog\/rediscover-redis-for-vector-similarity-search\/\" target=\"_blank\">Redis<\/a>\uff0c\u4f5c\u4e3a\u77e2\u91cf\u6570\u636e\u5e93\u7684&nbsp;<\/li>\n\n\n\n<li><a href=\"https:\/\/qdrant.tech\/\" target=\"_blank\" rel=\"noreferrer noopener\">Qdrant<\/a>\uff0c\u4e00\u4e2a\u77e2\u91cf\u641c\u7d22\u5f15\u64ce<\/li>\n\n\n\n<li><a href=\"https:\/\/typesense.org\/docs\/0.24.0\/api\/vector-search.html\" target=\"_blank\" rel=\"noreferrer noopener\">Typesense<\/a>\uff0c\u5f00\u6e90\u641c\u7d22\u5f15\u64ce\uff0c\u5e26\u77e2\u91cf\u641c\u7d22<\/li>\n<\/ul>\n\n\n\n<p><a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\/which-distance-function-should-i-use\"><\/a><\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><a href=\"https:\/\/platform.openai.com\/docs\/guides\/embeddings\/which-distance-function-should-i-use\">\u6211\u5e94\u8be5\u4f7f\u7528\u54ea\u79cd\u8ddd\u79bb\u51fd\u6570\uff1f<\/a><\/h3>\n\n\n\n<p>\u6211\u4eec\u5efa\u8bae<a href=\"https:\/\/en.wikipedia.org\/wiki\/Cosine_similarity\" target=\"_blank\" rel=\"noreferrer noopener\">\u4f59\u5f26\u76f8\u4f3c\u6027<\/a>\u3002\u8ddd\u79bb\u51fd\u6570\u7684\u9009\u62e9\u901a\u5e38\u5e76\u4e0d\u91cd\u8981\u3002<\/p>\n\n\n\n<p>OpenAI \u5d4c\u5165\u88ab\u89c4\u8303\u5316\u4e3a\u957f\u5ea6 1\uff0c\u8fd9\u610f\u5473\u7740\uff1a<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>\u4f59\u5f26\u76f8\u4f3c\u6027\u7684\u8ba1\u7b97\u901f\u5ea6\u7a0d\u5feb\uff0c\u53ea\u9700\u4f7f\u7528\u70b9\u79ef\u5373\u53ef<\/li>\n\n\n\n<li>\u4f59\u5f26\u76f8\u4f3c\u6027\u548c\u6b27\u51e0\u91cc\u5f97\u8ddd\u79bb\u5c06\u5bfc\u81f4\u76f8\u540c\u7684\u6392\u540d<\/li>\n<\/ul>\n","protected":false},"excerpt":{"rendered":"<p>\u4ec0\u4e48\u662f\u5d4c\u5165\uff1f OpenAI\u7684\u6587\u672c\u5d4c\u5165\u8861\u91cf\u6587\u672c\u5b57\u7b26\u4e32\u7684\u76f8\u5173\u6027\u3002\u5d4c\u5165\u901a\u5e38\u7528\u4e8e\uff1a \u539f\u6587\u94fe\u63a5\uff1aEmbeddings &#038; [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"site-sidebar-layout":"default","site-content-layout":"","ast-site-content-layout":"","site-content-style":"default","site-sidebar-style":"default","ast-global-header-display":"","ast-banner-title-visibility":"","ast-main-header-display":"","ast-hfb-above-header-display":"","ast-hfb-below-header-display":"","ast-hfb-mobile-header-display":"","site-post-title":"","ast-breadcrumbs-content":"","ast-featured-img":"","footer-sml-layout":"","theme-transparent-header-meta":"","adv-header-id-meta":"","stick-header-meta":"","header-above-stick-meta":"","header-main-stick-meta":"","header-below-stick-meta":"","astra-migrate-meta-layouts":"default","ast-page-background-enabled":"default","ast-page-background-meta":{"desktop":{"background-color":"var(--ast-global-color-4)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"ast-content-background-meta":{"desktop":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"_jetpack_memberships_contains_paid_content":false,"footnotes":""},"categories":[289,312,43],"tags":[242,314],"class_list":["post-1320","post","type-post","status-publish","format-standard","hentry","category-gpt","category-openai","category-infoarticle","tag-chatgpt","tag-openai-api"],"views":1317,"jetpack_sharing_enabled":true,"jetpack_featured_media_url":"","_links":{"self":[{"href":"https:\/\/www.aqwu.net\/wp\/index.php?rest_route=\/wp\/v2\/posts\/1320","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.aqwu.net\/wp\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.aqwu.net\/wp\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.aqwu.net\/wp\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.aqwu.net\/wp\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=1320"}],"version-history":[{"count":2,"href":"https:\/\/www.aqwu.net\/wp\/index.php?rest_route=\/wp\/v2\/posts\/1320\/revisions"}],"predecessor-version":[{"id":1329,"href":"https:\/\/www.aqwu.net\/wp\/index.php?rest_route=\/wp\/v2\/posts\/1320\/revisions\/1329"}],"wp:attachment":[{"href":"https:\/\/www.aqwu.net\/wp\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=1320"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.aqwu.net\/wp\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=1320"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.aqwu.net\/wp\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=1320"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}