{"id":1932,"date":"2023-10-19T14:20:12","date_gmt":"2023-10-19T06:20:12","guid":{"rendered":"http:\/\/i-services.info\/cspclm\/?p=1932"},"modified":"2023-10-20T13:30:18","modified_gmt":"2023-10-20T05:30:18","slug":"azure-openai-%e6%a8%a1%e5%9e%8b%e6%af%94%e8%bc%83-gpt-3-5-turbo-vs-gpt-3-5-turbo-%e6%8c%87%e7%a4%ba","status":"publish","type":"post","link":"http:\/\/i-services.info\/cspclm\/?p=1932","title":{"rendered":"Azure OpenAI \u6a21\u578b\u6bd4\u8f03: GPT-3.5 Turbo vs. GPT-3.5 Turbo \u6307\u793a"},"content":{"rendered":"<p><font color=\"#000000\">\u5fae\u8edf\u4e5d\u6708\u516c\u5e03\u7684 Azure OpenAI \u6700\u65b0\u6d88\u606f\u4e2d\u63d0\u5230: <em>Azure OpenAI<\/em><em>\u670d\u52d9\u73fe\u5728\u652f\u63f4 GPT-3.5 Turbo \u6307\u793a\u6a21\u578b\u3002<\/em><\/font><\/p>\n<p><font color=\"#000000\">\u770b\u5230\u9019\u500b\u6d88\u606f\uff0c\u4e0d\u77e5\u9053\u5927\u5bb6\u6709\u6c92\u6709\u9583\u904e\u9019\u6a23\u7684\u60f3\u6cd5: <strong>GPT-3.5 Turbo <\/strong><strong>\u6307\u793a?<\/strong> \u96e3\u9053\u8ddf <strong>GPT-3.5 Turbo<\/strong> \u662f\u4e0d\u4e00\u6a23\u7684\u6a21\u578b\u55ce?\u5b83\u5011\u4e4b\u9593\u6709\u751a\u9ebc\u95dc\u806f\u5462\uff1f<\/font><\/p>\n<p><font color=\"#000000\">\u4eca\u5929\u9019\u7bc7\u6587\u7ae0\u5c31\u662f\u8981\u4f86\u4ecb\u7d39\u9019\u5169\u500b\u6a21\u578b\u7684\u5dee\u5225\uff0c\u642d\u914d\u5be6\u969b\u4f7f\u7528\u6848\u4f8b\uff0c\u6211\u5011\u5c31\u53ef\u4ee5\u66f4\u6e05\u695a\u5728\u751a\u9ebc\u6a23\u60c5\u5883\u4e0b\uff0c\u66f4\u9069\u5408\u9078\u64c7\u4f7f\u7528\u54ea\u500b\u6a21\u578b\u3002<\/font><\/p>\n<p><font color=\"#000000\">\u60f3\u77e5\u9053\u66f4\u591a <strong>Azure OpenAI <\/strong><strong>\u670d\u52d9<\/strong> \u7684\u6700\u65b0\u6d88\u606f\u53ef\u4ee5\u5728\u9019\u88e1\u627e\u5230:<\/font><\/p>\n<p><a href=\"https:\/\/learn.microsoft.com\/zh-tw\/azure\/ai-services\/openai\/whats-new\"><font color=\"#0000ff\">Azure OpenAI \u670d\u52d9\u6709\u54ea\u4e9b\u65b0\u529f\u80fd\uff1f &#8211; Azure AI services<\/font><\/a><\/p>\n<h2><font color=\"#000000\">\u751a\u9ebc\u662f \u300c<strong>GPT-3.5 Turbo <\/strong><strong>\u6307\u793a\u300d<\/strong><strong>?<\/strong><\/font><\/h2>\n<p><font color=\"#000000\">GPT-3.5 Turbo \u6307\u793a\uff0c\u82f1\u6587\u70ba\u300c<strong>GPT-3.5 Turbo Instruct<\/strong>\u300d\uff0c\u6b64\u6a21\u578b\u985e\u4f3c\u65bc\u300c<strong>text-davinci-003<\/strong><strong>\u300d<\/strong>\uff0c\u4f7f\u7528\u7684\u662f<strong>\u5b8c\u6210 API (Completion)<\/strong>\u3002<\/font><\/p>\n<p><font color=\"#000000\">\u622a\u81f3\u4eca\u65e5(2023\/10\/16)\uff0c\u76ee\u524d\u53ea\u6709\u5169\u500b\u5340\u57df\u80fd\u5efa\u7acb\u300cGPT-3.5 Turbo \u6307\u793a\u300d: \u7f8e\u570b\u6771\u90e8(East US), \u3001\u745e\u5178\u4e2d\u90e8(Sweden Central)\uff0c\u6a21\u578b\u7684\u6b0a\u6756\u4e0a\u9650(max token)\u70ba 4097 \u500b tokens\u3002<\/font><\/p>\n<p><font color=\"#000000\">\u4f46\u5728\u516d\u6708\u6642\uff0cAzure\u5e73\u53f0\u4e0a\u67b6GPT-35 Turbo \u548cGPT-4\u7b49\u4f7f\u7528<strong>\u804a\u5929\u5b8c\u6210(ChatCompletion)<\/strong> API\u7684\u6a21\u578b\u4f86\u53d6\u4ee3\u4f7f\u7528<strong>\u5b8c\u6210(Completion) API<\/strong>\u820a\u6709\u7684\u6a21\u578b\u3002<\/font><\/p>\n<p><font color=\"#000000\">\u9019\u88e1\u6240\u63d0\u5230\u7684\u300c<strong>text-davinci-003<\/strong>\u300d\u4e5f\u662f\u5176\u4e2d\u4e4b\u4e00\uff0c\u8a31\u591a\u57fa\u65bc\u6b64\u6a21\u578b\u6240\u958b\u767c\u7684\u61c9\u7528\u7a0b\u5f0f\u4e5f\u56e0\u6b64\u53d7\u5230\u5f71\u97ff\uff0c\u7a0b\u5f0f\u78bc\u4e5f\u5fc5\u9808\u6839\u64da\u300c\u804a\u5929\u5b8c\u6210API\u300d \u9032\u884c\u4fee\u6539\u3002<\/font><\/p>\n<p><\/p>\n<h2><font color=\"#000000\" style=\"font-weight: bold;\">\u5b8c\u6210(Completion)?\u804a\u5929\u5b8c\u6210(ChatCompletion) ?<\/font><\/h2>\n<p><font color=\"#000000\">\u5982\u679c\u6211\u5011\u60f3\u8981 Azure OpenAI \u6a21\u578b\u8f38\u51fa\u7684\u5167\u5bb9\u66f4\u7b26\u5408\u9810\u671f\u7684\u7d50\u679c\uff0c\u90a3\u6211\u5011\u5c31\u4e00\u5b9a\u8981\u77ad\u89e3<strong>\u5b8c\u6210(Completion)\u548c\u804a\u5929\u5b8c\u6210(ChatCompletion)<\/strong> \u9019\u5169\u500b API \u7684\u4e0d\u540c\u4e4b\u8655\u3002<\/font><\/p>\n<p><font color=\"#000000\">\u55ae\u7d14\u7684\u4efb\u52d9\u5728\u4f7f\u7528 <strong>Completion<\/strong> \u548c <strong>ChatCompletion<\/strong> \u5169\u7a2eAPI\u6642\u90fd\u53ef\u4ee5\u9054\u5230\u76f8\u8fd1\u7684\u7d50\u679c\uff0c\u4f46\u4f7f\u7528<strong>ChatCompletion<\/strong>\uff0c\u5728\u69cb\u5efa\u63d0\u793a(Prompt)\u6642\u8f03\u5177\u9748\u6d3b\u6027\u3002<\/font><\/p>\n<p><font color=\"#000000\">\u9019\u500b\u6a21\u578b\u70ba\u804a\u5929\u60c5\u5883\u9032\u884c\u4e86\u512a\u5316\uff0c\u5728\u529f\u80fd\u4e0a\uff0c<strong>ChatCompletion<\/strong> \u6709\u5b9a\u7fa9AI\u6a21\u578b\u7684\u7cfb\u7d71\u8a0a\u606f(System Message)\u9078\u9805\uff0c\u6b64\u5916\u9084\u5167\u5efa\u4e86\u63d0\u4f9b\u5148\u524d\u6d88\u606f\u7684\u7d50\u69cb\uff0c\u7576\u7136\u53ea\u8981\u5728\u7cfb\u7d71\u6d88\u606f\u4e2d\u8abf\u6574\u63d0\u793a\uff0c<strong>ChatCompletion<\/strong>\u4e5f\u53ef\u4ee5\u7528\u65bc\u975e\u804a\u5929\u5834\u666f\u3002<\/font><\/p>\n<p><font color=\"#000000\">\u4f7f\u7528<strong>Completion API<\/strong> \u4e5f\u53ef\u4ee5\u9054\u5230\u985e\u4f3c\u7684\u7d50\u679c\uff0c\u4f46\u5fc5\u9808\u984d\u5916\u52a0\u4e0a\u683c\u5f0f\u5316\u63d0\u793a\uff0c\u9808\u901a\u904e\u6240\u8b02\u7684\u5143<em>\u63d0\u793a(meta prompt)<\/em>\u4f86\u5be6\u73fe\uff0c\u4ee5\u4fbfAI\u6a21\u578b\u7406\u89e3\u524d\u5f8c\u63d0\u793a\u9593\u7684\u95dc\u4fc2\u3002<\/font><\/p>\n<p><font color=\"#000000\"><br \/><\/font><\/p>\n<h3><font color=\"#000000\" style=\"font-weight: bold;\">\u5b8c\u6210(Completion) vs. \u804a\u5929\u5b8c\u6210(ChatCompletion) \u6bd4\u8f03\u8868\u683c<\/font><\/h3>\n<table style=\"width: 897px; height: 336px;\" border=\"1\" cellspacing=\"0\" cellpadding=\"0\">\n<tbody>\n<tr>\n<td width=\"186\">\n<p><b><font color=\"#000000\">\u985e\u5225<\/font><\/b><\/p>\n<\/td>\n<td width=\"186\">\n<p><font color=\"#000000\"><b>\u5b8c\u6210\uff08Completion<\/b><b>\uff09<\/b><\/font><\/p>\n<\/td>\n<td width=\"186\">\n<p><font color=\"#000000\"><b>\u804a\u5929\u5b8c\u6210\uff08ChatCompletion<\/b><b>\uff09<\/b><\/font><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td width=\"186\">\n<p><font color=\"#000000\">\u5c0d\u8a71\u5f62\u5f0f<\/font><\/p>\n<\/td>\n<td width=\"186\">\n<p><font color=\"#000000\">\u55ae\u8f49\u5c0d\u8a71(Single-Turn)\uff0c\u7cfb\u7d71\u6839\u64da\u55ae\u500b\u8f38\u5165\u63d0\u793a\u751f\u6210\u56de\u61c9\u3002<\/font><\/p>\n<\/td>\n<td width=\"186\">\n<p><font color=\"#000000\">\u591a\u8f49\u5c0d\u8a71(Multi-Turn)\uff0c\u7cfb\u7d71\u6703\u8a18\u9304\u5c0d\u8a71\u6b77\u53f2\u4e26\u751f\u6210\u56de\u61c9\u3002<\/font><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td width=\"186\">\n<p><font color=\"#000000\">\u6a21\u578b<\/font><\/p>\n<\/td>\n<td width=\"186\">\n<p><font color=\"#000000\">GPT-3.5 Turbo Instruct (\u524d text-davinci-003)<\/font><\/p>\n<\/td>\n<td width=\"186\">\n<p><font color=\"#000000\">GPT-3.5 Turbo GPT-4<\/font><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td width=\"186\">\n<p><font color=\"#000000\">\u7528\u9014<\/font><\/p>\n<\/td>\n<td width=\"186\">\n<p><font color=\"#000000\">\u66f4\u9069\u5408\u55ae\u500b\u63d0\u793a\u4efb\u52d9\u3002<\/font><\/p>\n<\/td>\n<td width=\"186\">\n<p><font color=\"#000000\">\u9069\u7528\u65bc\u5177\u4e92\u52d5\u6027\u7684\u5c0d\u8a71\u61c9\u7528\u3002<\/font><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td width=\"186\">\n<p><font color=\"#000000\">API\u8a2d\u7f6e<\/font><\/p>\n<\/td>\n<td width=\"186\">\n<p><code><font color=\"#000000\">openai.Completion.create<\/font><\/code><\/p>\n<\/td>\n<td width=\"186\">\n<p><code><font color=\"#000000\">openai.ChatCompletion.create<\/font><\/code><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td width=\"186\">\n<p><font color=\"#000000\">\u9032\u968e\u8a2d\u7f6e<\/font><\/p>\n<\/td>\n<td width=\"186\">\n<p><font color=\"#000000\">\u63d0\u793a\u8a2d\u8a08\u548c\u5de5\u7a0b (prompt design and engineering)<\/font><\/p>\n<\/td>\n<td width=\"186\">\n<p><font color=\"#000000\">\u7cfb\u7d71\u8a0a\u606f\u53ca\u6a21\u677f\u8a2d\u8a08 (System message and System message template)<\/font><\/p>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<h2><font color=\"#000000\">Azure OpenAI Studio Playground<\/font><\/h2>\n<p><font color=\"#000000\">\u807d\u8d77\u4f86\u597d\u50cf\u9084\u662f\u6709\u9ede\u6a21\u7cca? \u6700\u597d\u7684\u65b9\u6cd5\u5c31\u662f\u76f4\u63a5\u5230 <\/font><a href=\"https:\/\/oai.azure.com\/portal\"><font color=\"#000000\">Azure OpenAI Studio<\/font><\/a><font color=\"#000000\"> \u4f86\u89aa\u81ea\u5617\u8a66\u5169\u500b\u6a21\u578b:<\/font><\/p>\n<p><font color=\"#000000\">\u5148\u6c7a\u689d\u4ef6:<\/font><\/p>\n<ul>\n<li><font color=\"#000000\">Azure \u8a02\u95b1<\/font>\n<li><font color=\"#000000\">Azure OpenAI Service \u8cc7\u6e90<\/font>\n<li><font color=\"#000000\">\u90e8\u7f72\u300cGPT-3.5 Turbo\u300d \u6a21\u578b<\/font>\n<li><font color=\"#000000\">\u90e8\u7f72\u300cGPT-3.5 Turbo Instruct \u300d\u6a21\u578b<\/font><\/li>\n<\/ul>\n<p><strong><font color=\"#000000\">\u5b8c\u6210(Completion)<\/font><\/strong><\/p>\n<p><strong><a href=\"http:\/\/i-services.info\/cspclm\/wp-content\/uploads\/2023\/10\/Untitled.png\"><font color=\"#000000\"><img loading=\"lazy\" width=\"854\" height=\"426\" title=\"Untitled\" style=\"display: inline; background-image: none;\" alt=\"Untitled\" src=\"http:\/\/i-services.info\/cspclm\/wp-content\/uploads\/2023\/10\/Untitled_thumb.png\" border=\"0\"><\/font><\/a><\/strong><\/p>\n<p><strong><font color=\"#000000\">\u804a\u5929\u5b8c\u6210(ChatCompletion)<\/font><\/strong><\/p>\n<p><strong><a href=\"http:\/\/i-services.info\/cspclm\/wp-content\/uploads\/2023\/10\/Untitled-1.png\"><font color=\"#000000\"><img loading=\"lazy\" width=\"854\" height=\"436\" title=\"Untitled 1\" style=\"display: inline; background-image: none;\" alt=\"Untitled 1\" src=\"http:\/\/i-services.info\/cspclm\/wp-content\/uploads\/2023\/10\/Untitled-1_thumb.png\" border=\"0\"><\/font><\/a><\/strong><\/p>\n<p><font color=\"#000000\">\u6309\u4e0b\u300c<strong>View code\u300d<\/strong>\u6211\u5011\u53ef\u4ee5\u770b\u51fa\uff0c\u5169\u8005\u4e4b\u9593\u7a0b\u5f0f\u78bc\u4e5f\u6709\u5340\u5225\uff1a<\/font><\/p>\n<p><font color=\"#000000\">\u5b8c\u6210(Completion) \u7a0b\u5f0f\u78bc:<\/font><\/p>\n<ol>\n<li><code><em><font color=\"#000000\">openai.Completion.create<\/font><\/em><\/code>\n<li><font color=\"#000000\">\u4f7f\u7528\u8005\u4ee5Prompt(\u63d0\u793a)\u4f86\u544a\u8a34\u6a21\u578b\u8981\u5b8c\u6210(Completion)\u751a\u9ebc\u7d50\u679c<\/font><\/li>\n<\/ol>\n<p><a href=\"http:\/\/i-services.info\/cspclm\/wp-content\/uploads\/2023\/10\/Untitled-2.png\"><font color=\"#000000\"><img loading=\"lazy\" width=\"854\" height=\"749\" title=\"Untitled 2\" style=\"display: inline; background-image: none;\" alt=\"Untitled 2\" src=\"http:\/\/i-services.info\/cspclm\/wp-content\/uploads\/2023\/10\/Untitled-2_thumb.png\" border=\"0\"><\/font><\/a><\/p>\n<p><font color=\"#000000\">\u804a\u5929\u5b8c\u6210(ChatCompletion)\u7a0b\u5f0f\u78bc:<\/font><\/p>\n<ol>\n<li><code><em><font color=\"#000000\">openai.ChatCompletion.create<\/font><\/em><\/code>\n<li><font color=\"#000000\">\u8ce6\u4e88\u6a21\u578b\u4e00\u500b\u89d2\u8272(system)<\/font>\n<li><font color=\"#000000\">\u4f7f\u7528\u8005(user)\u8207\u6a21\u578b(assistant)\u900f\u904e\u804a\u5929\u5f62\u5f0f\u9032\u884c\u6e9d\u901a<\/font><\/li>\n<\/ol>\n<p><strong><a href=\"http:\/\/i-services.info\/cspclm\/wp-content\/uploads\/2023\/10\/Untitled-3.png\"><font color=\"#000000\"><img loading=\"lazy\" width=\"854\" height=\"688\" title=\"Untitled 3\" style=\"display: inline; background-image: none;\" alt=\"Untitled 3\" src=\"http:\/\/i-services.info\/cspclm\/wp-content\/uploads\/2023\/10\/Untitled-3_thumb.png\" border=\"0\"><\/font><\/a><\/strong><\/p>\n<h2><font color=\"#000000\" style=\"font-weight: bold;\">GPT-3.5 Turbo vs. GPT-3.5 Turbo \u6307\u793a \u6a21\u578b\u8868\u73fe<\/font><\/h2>\n<p><font color=\"#000000\">\u77ad\u89e3\u4e86\u5169\u500b\u6a21\u578b\u7684\u57fa\u672c\u6982\u5ff5\u4ee5\u53ca\u5dee\u5225\u5f8c\uff0c\u63a5\u4e0b\u4f86\u6211\u5011\u8b93\u5169\u500b\u6a21\u578b\u57f7\u884c5\u500b\u4efb\u52d9\uff0c\u4e26\u7528\u4e09\u500b\u6e2c\u91cf\u6a19\u6e96\u4f86\u5224\u65b7\u6a21\u578b\u7684\u8868\u73fe\uff0c\u5206\u5225\u662f:<\/font><\/p>\n<ul>\n<li><font color=\"#000000\"><strong><code>CPU times<\/code><\/strong>: \u7a0b\u5f0f\u5728\u4e2d\u592e\u8655\u7406\u5668\uff08CPU\uff09\u4e0a\u904b\u884c\u7684\u7e3d\u6642\u9593\u3002<br \/>\n<\/font><\/p>\n<li><font color=\"#000000\"><strong><code>Wall time<\/code><\/strong>:\u958b\u59cb\u57f7\u884c\u7a0b\u5f0f\u5230\u7a0b\u5f0f\u5b8c\u6210\u7684\u5be6\u969b\u7d93\u904e\u7684\u6642\u9593\u3002\u5305\u62ecI\/O\u64cd\u4f5c\u3001\u7b49\u5f85\u5176\u4ed6\u7a0b\u5e8f\u3001\u7cfb\u7d71\u8abf\u5ea6\u7b49\u6642\u9593\u3002<br \/>\n<\/font><\/p>\n<li><font color=\"#000000\"><code><strong>\u6a21\u578b\u5b8c\u6210(\u56de\u8986)<\/strong><\/code>: \u6a21\u578b\u6240\u56de\u8986\u7684\u5b8c\u6574\u5b57\u4e32\u3002 <\/font><\/li>\n<\/ul>\n<p><font color=\"#000000\"><br \/><\/font><\/p>\n<h3><font color=\"#000000\" style=\"font-weight: bold;\">\u4efb\u52d9#1\uff1a\u662f\u975e\u984c<\/font><\/h3>\n<p><font color=\"#000000\"><br \/>\n<\/font><\/p>\n<h4><font color=\"#000000\">\u984c\u76ee(Prompt): Microsoft \u6709\u6536\u8cfcOpenAI \u55ce?<\/font><\/h4>\n<table style=\"width: 968px; height: 516px;\" border=\"1\" cellspacing=\"0\" cellpadding=\"0\">\n<tbody>\n<tr>\n<td width=\"139\">\n<p><font color=\"#000000\">\u6a21\u578b\u540d\u7a31<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">CPU times<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">Wall time:<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">(Chat)Completion<\/font><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td width=\"139\">\n<p><font color=\"#000000\">gpt-35-turbo<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">31.2 ms<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">2.1 s<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">Microsoft \u6c92\u6709\u5b8c\u5168\u6536\u8cfc OpenAI\uff0c\u4f46\u5728 2019 \u5e74\uff0cMicrosoft \u548c OpenAI \u7c3d\u7f72\u4e86\u4e00\u4efd\u5408\u4f5c\u5354\u8b70\uff0c\u96d9\u65b9\u5c07\u5171\u540c\u958b\u767c\u4eba\u5de5\u667a\u6167\u6280\u8853\uff0c\u4e26\u5c07 Microsoft Azure \u4f5c\u70ba OpenAI \u7684\u9996\u9078\u96f2\u5e73\u53f0\u3002\u6b64\u5916\uff0cMicrosoft \u4e5f\u6210\u70ba\u4e86 OpenAI \u7684\u7368\u5bb6\u96f2\u670d\u52d9\u63d0\u4f9b\u5546\u3002\u9019\u4efd\u5408\u4f5c\u5354\u8b70\u8b93 Microsoft \u80fd\u5920\u7372\u5f97 OpenAI \u7684\u6280\u8853\u548c\u5c08\u696d\u77e5\u8b58\uff0c\u4e26\u5c07\u5176\u6574\u5408\u5230\u81ea\u5df1\u7684\u7522\u54c1\u548c\u670d\u52d9\u4e2d\u3002<\/font><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td width=\"139\">\n<p><font color=\"#000000\">gpt-35-turbo-instruct<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">0 ns <\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">1.74 s<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">\u76ee\u524d\u5c1a\u672a\u6709\u4efb\u4f55\u6d88\u606f\u8868\u660eMicrosoft\u6709\u6536\u8cfcOpenAI\u7684\u8a08\u5283\u3002OpenAI\u662f\u4e00\u5bb6\u7368\u7acb\u7684\u4eba\u5de5\u667a\u80fd\u7814\u7a76\u6a5f\u69cb\uff0c\u5176\u5275\u59cb\u4eba\u5305\u62ec\u57c3\u9686\u00b7\u99ac\u65af\u514b\u548c\u5176\u4ed6\u77e5\u540d\u4f01\u696d\u5bb6\uff0c\u4e26\u4e14\u8207\u591a\u5bb6\u516c\u53f8\u5408\u4f5c\uff0c\u5305\u62ecMicrosoft\u3002\u4f46\u662f\uff0cMicrosoft\u4e26\u672a\u516c\u958b\u8868\u793a\u6709\u6536\u8cfcOpenAI\u7684\u610f\u5411\u3002<\/font><\/p>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p><p><font color=\"#000000\"><br \/><\/font><\/p>\n<p><strong><font color=\"#000000\">\u4efb\u52d9#2\uff1a\u540c\u60c5\u5fc3<\/font><\/strong><\/p>\n<h4><font color=\"#000000\">\u984c\u76ee(Prompt): \u6211\u597d\u96e3\u904e\uff0c\u4f60\u53ef\u4ee5\u8aaa\u4e9b\u8a71\u9f13\u52f5\u6211\u55ce?<\/font><\/h4>\n<table style=\"width: 970px; height: 426px;\" border=\"1\" cellspacing=\"0\" cellpadding=\"0\">\n<tbody>\n<tr>\n<td width=\"139\">\n<p><font color=\"#000000\">\u6a21\u578b\u540d\u7a31<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">CPU times<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">Wall time:<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">(Chat)Completion<\/font><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td width=\"139\">\n<p><font color=\"#000000\">gpt-35-turbo<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">15.6 ms<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">3.44 s<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">\u7576\u4f60\u611f\u5230\u96e3\u904e\u6642\uff0c\u8acb\u8a18\u4f4f\u4f60\u4e0d\u662f\u5b64\u55ae\u7684\u3002\u4f60\u6709\u5bb6\u4eba\u3001\u670b\u53cb\u548c\u652f\u6301\u4f60\u7684\u4eba\u3002\u4e5f\u8acb\u8a18\u4f4f\uff0c\u9019\u7a2e\u60c5\u7dd2\u662f\u66ab\u6642\u7684\uff0c\u4f60\u6703\u5ea6\u904e\u9019\u500b\u96e3\u95dc\u3002\u8acb\u76f8\u4fe1\u81ea\u5df1\uff0c\u4f60\u6709\u8db3\u5920\u7684\u529b\u91cf\u548c\u52c7\u6c23\u53bb\u514b\u670d\u4efb\u4f55\u56f0\u96e3\u3002\u4e0d\u8981\u5bb3\u6015\u5411\u4ed6\u4eba\u5c0b\u6c42\u5e6b\u52a9\uff0c\u4ed6\u5011\u6703\u5f88\u6a02\u610f\u5e6b\u52a9\u4f60\u3002\u6700\u91cd\u8981\u7684\u662f\uff0c\u8acb\u4fdd\u6301\u7a4d\u6975\u7684\u614b\u5ea6\u548c\u4fe1\u5fc3\uff0c\u76f8\u4fe1\u81ea\u5df1\u53ef\u4ee5\u514b\u670d\u4e00\u5207\u3002<\/font><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td width=\"139\">\n<p><font color=\"#000000\">gpt-35-turbo-instruct<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">0 ns<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">2.19 s<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">\u7576\u7136\u53ef\u4ee5\uff01\u9996\u5148\uff0c\u6211\u60f3\u544a\u8a34\u4f60\uff0c\u6bcf\u500b\u4eba\u90fd\u6703\u9047\u5230\u56f0\u96e3\u548c\u6311\u6230\uff0c\u9019\u662f\u751f\u6d3b\u4e2d\u4e0d\u53ef\u907f\u514d\u7684\u4e00\u90e8\u5206\u3002\u4f46\u662f\uff0c\u91cd\u8981\u7684\u662f\u8981\u8a18\u4f4f\uff0c\u4f60\u4e26\u4e0d\u5b64\u55ae\uff0c\u6211\u5011\u90fd\u5728\u9019\u88e1\u652f\u6301\u4f60\u3002\u4f60\u53ef\u4ee5\u76f8\u4fe1\u81ea\u5df1\u7684\u80fd\u529b\uff0c\u52c7\u6562\u5730\u9762\u5c0d\u56f0\u96e3\uff0c\u56e0\u70ba\u4f60\u6709\u8db3\u5920\u7684\u529b\u91cf\u548c\u667a\u6167\u53bb\u514b\u670d\u5b83\u5011\u3002 \u6b64\u5916\uff0c\u8acb\u8a18\u4f4f\uff0c\u6bcf\u4e00\u6b21\u632b\u6298\u90fd\u662f\u4e00\u6b21\u6210\u9577\u7684\u6a5f\u6703\u3002\u5b83\u53ef\u4ee5\u5e6b\u52a9\u4f60\u5b78\u7fd2\u66f4\u591a\uff0c\u6210\u70ba\u4e00\u500b\u66f4\u5f37\u5927\u7684\u4eba\u3002\u4e0d\u8981\u5bb3\u6015\u72af\u932f\uff0c\u56e0\u70ba\u5b83\u5011\u53ef\u4ee5\u5e6b\u52a9\u4f60\u6210\u9577\u548c\u9032\u6b65\u3002 \u6700\u5f8c\uff0c\u8acb\u76f8\u4fe1\u81ea\u5df1\uff0c\u76f8\u4fe1\u4f60\u7684\u5922\u60f3\u548c\u76ee\u6a19<\/font><\/p>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p><font color=\"#000000\"><br \/><\/font><\/p>\n<h3><font color=\"#000000\" style=\"font-weight: bold;\">\u4efb\u52d9#3\uff1a\u7a0b\u5f0f\u78bc\u751f\u6210<\/font><\/h3>\n<h4><font color=\"#000000\">\u984c\u76ee(Prompt): \u53ef\u4ee5\u8acb\u4f60\u5e6b\u6211\u5beb\u500bPython \u51fd\u5f0f, \u8b93\u6211\u80fd\u8b80\u53d6csv \u6a94\u6848, \u4e26\u56de\u50b3\u6bcf\u5169\u884c\u7684\u5167\u5bb9\u55ce?<\/font><\/h4>\n<table style=\"width: 983px; height: 1113px;\" border=\"1\" cellspacing=\"0\" cellpadding=\"0\">\n<tbody>\n<tr>\n<td width=\"139\">\n<p><font color=\"#000000\">\u6a21\u578b\u540d\u7a31<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">CPU times<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">Wall time:<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">(Chat)Completion<\/font><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td width=\"139\">\n<p><font color=\"#000000\">gpt-35-turbo<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">0 ns<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">4.48 s<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">\u7576\u7136\u53ef\u4ee5\uff0c\u4ee5\u4e0b\u662f\u4e00\u500b\u53ef\u4ee5\u8b80\u53d6csv\u6a94\u6848\u4e26\u56de\u50b3\u6bcf\u5169\u884c\u5167\u5bb9\u7684Python\u51fd\u5f0f\uff1a <code>python import csv def read_csv_file(file_path): with open(file_path, 'r') as csv_file: csv_reader = csv.reader(csv_file) rows = [] for row in csv_reader: rows.append(row) return [rows[i:i+2] for i in range(0, len(rows), 2)] <\/code>\u9019\u500b\u51fd\u5f0f\u4f7f\u7528Python\u7684csv\u6a21\u7d44\u4f86\u8b80\u53d6csv\u6a94\u6848\uff0c\u4e26\u5c07\u6bcf\u4e00\u884c\u7684\u5167\u5bb9\u5b58\u5132\u5728\u4e00\u500b\u5217\u8868\u4e2d\u3002\u7136\u5f8c\uff0c\u4f7f\u7528\u5217\u8868\u63a8\u5c0e\u5f0f\u4f86\u5c07\u76f8\u9130\u7684\u5169\u884c\u7d44\u5408\u6210\u4e00\u500b\u5b50\u5217\u8868\uff0c\u6700\u5f8c\u8fd4\u56de\u4e00\u500b\u5305\u542b\u6240\u6709\u5b50\u5217\u8868\u7684\u5217\u8868\u3002 \u4f60\u53ef\u4ee5\u4f7f\u7528\u4ee5\u4e0b\u4ee3\u78bc\u4f86\u6e2c\u8a66\u9019\u500b\u51fd\u5f0f\uff1a <code>python file_path = 'example.csv' result = read_csv_file(file_path) print(result) <\/code>\u9019\u500b\u4ee3\u78bc\u6703\u8b80\u53d6\u540d\u70baexample.csv\u7684\u6a94\u6848\uff0c\u4e26\u5c07\u6bcf\u5169\u884c\u7684\u5167\u5bb9\u7d44\u5408\u6210\u4e00\u500b\u5b50\u5217\u8868\uff0c\u6700\u5f8c\u8f38\u51fa\u7d50\u679c\u3002\u4f60\u53ef\u4ee5\u6839\u64da\u81ea\u5df1\u7684\u9700\u6c42\u4fee\u6539\u9019\u500b\u51fd\u5f0f\uff0c\u4f8b\u5982\u66f4\u6539\u6bcf\u5169\u884c\u8b80\u53d6\u7684\u65b9\u5f0f\u6216\u8655\u7406\u6bcf\u884c\u7684\u5167\u5bb9\u3002<\/font><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td width=\"139\">\n<p><font color=\"#000000\">gpt-35-turbo-instruct<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">0 ns<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">1.62 s<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">def read_csv(filename): # \u958b\u555f\u6a94\u6848 file = open(filename, &#8216;r&#8217;) # \u8b80\u53d6\u6a94\u6848\u5167\u5bb9 lines = file.readlines() # \u95dc\u9589\u6a94\u6848 file.close() # \u5efa\u7acb\u7a7a\u7684list\u4f86\u5b58\u653e\u6bcf\u5169\u884c\u7684\u5167\u5bb9 result = [] # \u4f7f\u7528\u8ff4\u5708\u4f86\u8b80\u53d6\u6bcf\u5169\u884c\u7684\u5167\u5bb9 for i in range(0, len(lines), 2): # \u5c07\u6bcf\u5169\u884c\u7684\u5167\u5bb9\u52a0\u5165result list\u4e2d result.append(lines[i].strip() + &#8216;,&#8217; + lines[i+1].strip()) # \u56de\u50b3\u7d50\u679c return result<\/font><\/p>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<h3><font color=\"#000000\"><br \/><\/font><\/h3>\n<h3><font color=\"#000000\" style=\"font-weight: bold;\">\u4efb\u52d9#4\uff1a\u60c5\u611f\u5206\u6790<\/font><\/h3>\n<h4><font color=\"#000000\">\u984c\u76ee(Prompt): \u8acb\u63d0\u4f9b\u4ee5\u4e0b\u6bb5\u843d\u7684\u60c5\u611f\u5206\u6790: &#8220;\u53ef\u53e3\u53ef\u6a02\u516c\u53f8\u7b2c\u4e09\u8ca1\u5b63\u80a1\u6771\u61c9\u4f54\u76c8\u522928.25\u5104\u7f8e\u5143,\u6309\u5e74\u589e\u957714%;\u7d93\u8abf\u6574\u5f8c\u6bcf\u80a1\u76c8\u522969\u7f8e\u4ed9,\u591a\u904e\u5e02\u5834\u9810\u671f\u3002\u201d<\/font><\/h4>\n<table style=\"width: 988px; height: 292px;\" border=\"1\" cellspacing=\"0\" cellpadding=\"0\">\n<tbody>\n<tr>\n<td width=\"139\">\n<p><font color=\"#000000\">\u6a21\u578b\u540d\u7a31<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">CPU times<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">Wall time:<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">(Chat)Completion<\/font><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td width=\"139\">\n<p><font color=\"#000000\">gpt-35-turbo<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">46.9 ms<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">1.16 s<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">\u6b63\u9762\u60c5\u611f\u3002<\/font><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td width=\"139\">\n<p><font color=\"#000000\">gpt-35-turbo-instruct<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">31.2 ms<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">2.24 s<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">\u9019\u6bb5\u8a71\u8868\u9054\u4e86\u53ef\u53e3\u53ef\u6a02\u516c\u53f8\u5728\u7b2c\u4e09\u8ca1\u5b63\u53d6\u5f97\u7684\u826f\u597d\u696d\u7e3e\uff0c\u80a1\u6771\u61c9\u4f54\u76c8\u5229\u9054\u523028.25\u5104\u7f8e\u5143\uff0c\u540c\u6bd4\u589e\u957714%\uff0c\u6bcf\u80a1\u76c8\u5229\u4e5f\u8d85\u904e\u5e02\u5834\u9810\u671f\uff0c\u9054\u523069\u7f8e\u4ed9\u3002\u5f9e\u60c5\u611f\u5206\u6790\u7684\u89d2\u5ea6\u4f86\u770b\uff0c\u9019\u6bb5\u8a71\u5e36\u6709\u7a4d\u6975\u7684\u60c5\u7dd2\uff0c\u8868\u660e\u53ef\u53e3\u53ef\u6a02\u516c\u53f8\u5728\u7b2c\u4e09\u8ca1\u5b63\u53d6\u5f97\u4e86\u4ee4\u4eba\u6eff\u610f\u7684\u6210\u7e3e\uff0c\u80a1\u6771\u61c9\u8a72\u6703\u611f\u5230\u9ad8\u8208\u548c\u6eff\u610f\u3002\u6b64\u5916\uff0c\u63d0\u5230\u6bcf\u80a1\u76c8\u5229\u8d85\u904e\u5e02\u5834\u9810\u671f\uff0c\u4e5f\u6697\u793a\u8457\u5e02\u5834\u5c0d\u53ef\u53e3\u53ef\u6a02\u516c\u53f8\u7684\u672a\u4f86\u8868\u73fe\u6301\u6709\u4fe1\u5fc3\uff0c\u9019\u4e5f\u6703\u8b93\u6295\u8cc7\u8005\u611f\u5230\u6a02\u89c0\u3002<\/font><\/p>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p><font color=\"#000000\"><br \/><\/font><\/p>\n<h3><font color=\"#000000\" style=\"font-weight: bold;\">\u4efb\u52d9#5\uff1a\u6458\u8981\u7dad\u57fa\u767e\u79d1\u9801\u9762<\/font><\/h3>\n<h4><font color=\"#000000\">\u984c\u76ee(Prompt): \u641c\u5c0b:\u53f0\u5317101\uff0c\u8acb\u544a\u8a34\u62115\u500b\u6709\u8da3\u7684\u4e8b\u5be6:<\/font><\/h4>\n<table style=\"width: 1003px; height: 1165px;\" border=\"1\" cellspacing=\"0\" cellpadding=\"0\">\n<tbody>\n<tr>\n<td width=\"139\">\n<p><font color=\"#000000\">\u6a21\u578b\u540d\u7a31<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">CPU times<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">Wall time:<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">(Chat)Completion<\/font><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td width=\"139\">\n<p><font color=\"#000000\">gpt-35-turbo<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">0 ns<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">3.82 s<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">\u5728COVID-19\u5927\u6d41\u884c\u671f\u9593\uff0c\u53f0\u5317101\u88ab\u7528\u4f5c\u5de8\u5927\u7684\u9ad4\u6eab\u8a08\u3002\u8a72\u5efa\u7bc9\u7684LED\u71c8\u6703\u6839\u64da\u6eab\u5ea6\u8b8a\u5316\u984f\u8272\uff0c\u85cd\u8272\u8868\u793a\u6eab\u5ea6\u4f4e\u65bc37.5\u00b0C (99.5\u00b0F)\uff0c\u5448\u73fe\u7d05\u8272\u5247\u8868\u793a\u6eab\u5ea6\u9ad8\u65bc37.5\u00b0C\u3002\u9019\u6709\u52a9\u65bc\u63d0\u9192\u4eba\u5011\u6e2c\u91cf\u4ed6\u5011\u7684\u9ad4\u6eab\u4e26\u8b66\u60d5\u75c5\u6bd2\u7684\u50b3\u64ad\u3002 \u53f0\u5317101\u66fe\u51fa\u73fe\u5728\u591a\u90e8\u96fb\u5f71\u548c\u96fb\u8996\u7bc0\u76ee\u4e2d\uff0c\u5305\u62ec\u96fb\u5f71&#8221;\u9732\u897f&#8221;\u548c&#8221;\u9a5a\u5947\u4e4b\u65c519&#8243;\u3002\u5b83\u9084\u662f\u591a\u90e8\u7d00\u9304\u7247\u7684\u4e3b\u984c\uff0c\u5305\u62ec&#8221;\u53f0\u5317101\uff1a\u5efa\u9020\u6700\u9ad8\u5854&#8221;\u548c&#8221;\u53f0\u5317101\uff1a\u4e16\u754c\u4e0a\u6700\u9ad8\u7684\u7da0\u8272\u5efa\u7bc9&#8221;\u3002\u8a72\u5efa\u7bc9\u7684\u7368\u7279\u8a2d\u8a08\u548c\u5de5\u7a0b\u4f7f\u5176\u6210\u70ba\u904a\u5ba2\u548c\u5efa\u7bc9\u611b\u597d\u8005\u7684\u71b1\u9580\u76ee\u7684\u5730\u3002 \u53f0\u5317101\u5171\u6709101\u5c64\uff0c\u7b2c91\u5230101\u5c64\u7531\u53f0\u5317101\u5929\u6587\u53f0\u4f54\u7528\u3002\u5929\u6587\u53f0\u63d0\u4f9b\u53f0\u5317\u53ca\u5176\u5468\u908a\u5730\u5340\u7684\u5168\u666f\uff0c\u662f\u53d7\u6b61\u8fce\u7684\u65c5\u904a\u666f\u9ede\u3002\u904a\u5ba2\u53ef\u4ee5\u4e58\u5750\u9ad8\u901f\u96fb\u68af\u524d\u5f80\u4f4d\u65bc\u7b2c91\u5c64\u7684\u5929\u6587\u53f0\uff0c\u8a72\u5929\u6587\u53f0\u63d0\u4f9b\u5ba4\u5167\u548c\u5ba4\u5916\u89c0\u666f\u5340\u3002 \u53f0\u5317101\u64c1\u6709\u591a\u9805\u4e16\u754c\u7d00\u9304\uff0c\u5305\u62ec\u4e16\u754c\u4e0a\u6700\u5feb\u7684\u96fb\u68af\u3001\u4e16\u754c\u4e0a\u6700\u9ad8\u7684\u5ba4\u5916\u89c0\u666f\u53f0\u548c\u4e16\u754c\u4e0a\u6700\u5927\u7684\u65b0\u5e74\u524d\u5915\u7159\u706b\u8868\u6f14\u3002\u53f0\u5317101\u7684\u8de8\u5e74\u7159\u706b\u8868\u6f14\u5df2\u6210\u70ba\u4e00\u500b\u91cd\u5927\u6d3b\u52d5\uff0c\u6210\u5343\u4e0a\u842c\u7684\u4eba\u805a\u96c6\u5728\u5468\u908a\u5730\u5340\u89c0\u770b\u8868\u6f14\u3002\u8a72\u8868\u6f14\u5305\u542b\u8d85\u904e30,000\u767c\u7159\u706b\uff0c\u6301\u7e8c\u5e7e\u5206\u9418\uff0c\u9ad8\u6f6e\u8fed\u8d77\uff0c\u7167\u4eae\u6574\u5ea7\u5efa\u7bc9\u3002<\/font><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td width=\"139\">\n<p><font color=\"#000000\">gpt-35-turbo-instruct<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">0 ns<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">1.52 s<\/font><\/p>\n<\/td>\n<td width=\"139\">\n<p><font color=\"#000000\">101\u662f\u53f0\u7063\u6700\u53d7\u6b61\u8fce\u7684\u65c5\u904a\u666f\u9ede\u3002 1. \u5f9e2004\u5e74\u958b\u5e55\u52302009\u5e74\uff0c\u53f0\u5317101\u662f\u4e16\u754c\u4e0a\u6700\u9ad8\u7684\u5efa\u7bc9\uff0c\u76f4\u5230\u88ab\u54c8\u5229\u767c\u5854\u8d85\u8d8a\u3002 2. \u53f0\u5317101\u7684\u96fb\u68af\u53ef\u4ee5\u5728\u77ed\u77ed37\u79d2\u5167\u5f9e\u7b2c5\u5c64\u79fb\u52d5\u5230\u7b2c89\u5c64\uff0c\u5275\u4e0b\u6700\u5feb\u96fb\u68af\u901f\u5ea6\u7d00\u9304\u3002 3. \u53f0\u5317101\u662f\u4e16\u754c\u4e0a\u6700\u9ad8\u548c\u6700\u5927\u7684\u7da0\u8272\u5efa\u7bc9\uff0c\u65bc2011\u5e74\u6839\u64daLEED\u8a8d\u8b49\u7cfb\u7d71\u7372\u5f97Platinum\u8b49\u66f8\u8a55\u7d1a\u3002 4. \u8a72\u5efa\u7bc9\u7684\u5f8c\u73fe\u4ee3\u5efa\u7bc9\u98a8\u683c\u7d50\u5408\u4e86\u50b3\u7d71\u7684\u4e9e\u6d32\u7f8e\u5b78\u548c\u73fe\u4ee3\u7684\u5de5\u696d\u6750\u6599\u3002 5. \u53f0\u5317101\u7531\u53f0\u5317\u91d1\u878d\u4e2d\u5fc3\u516c\u53f8\u64c1\u6709\uff0c\u4e26\u8a08\u5283\u572870\u5e74\u79df\u7d04\u7d50\u675f\u5f8c\u65bc2067\u5e74\u6b78\u9084\u7d66\u53f0\u5317\u5e02\u3002<\/font><\/p>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>\u900f\u904e\u4e0a\u9762\u7684\u4efb\u52d9\u6211\u5011\u53ef\u4ee5\u770b\u5230\uff0c\u300cGPT-3.5 Turbo\u300d\u5728CPU\u904b\u7b97\u4ee5\u53ca\u7a0b\u5f0f\u57f7\u884c\u7684\u6642\u9593\u90fd\u6703\u8f03\u300cGPT-3.5 Turbo \u6307\u793a\u300d\u4f86\u5f97\u591a\u4e00\u4e9b\uff0c\u81f3\u65bc\u5728\u56de\u8986\u7684\u90e8\u5206\u5247\u662f\u5404\u6709\u5343\u79cb\uff0c\u6709\u5404\u81ea\u7684\u4eae\u9ede\uff0c\u8b80\u8005\u6bd4\u8f03\u559c\u6b61\u54ea\u4e00\u7a2e\u5462\uff1f<\/p>\n<h2>\n<h2><font color=\"#000000\" style=\"font-weight: bold;\">\u7d50\u8ad6<\/font><\/h2>\n<\/h2>\n<p>\u4eca\u5929\u8a0e\u8ad6\u5fae\u8edfAzure OpenAI\u4e2d\u7684\u5169\u500b\u5927\u578b\u8a9e\u8a00\u6a21\u578b\uff1a \u300cGPT-3.5 Turbo\u300d\u548c\u300cGPT-3.5 Turbo \u6307\u793a\u300d\uff0c\u96d6\u7136\u540d\u5b57\u53ea\u5dee\u4e00\u9ede\uff0c\u4f7f\u7528\u7684\u65b9\u5f0f\u9084\u662f\u6709\u4e9b\u8a31\u5dee\u5225\u3002<\/p>\n<p>\u800c\u4e0a\u8ff0\u63a2\u8a0e\u7684\u300cCompletion API\u300d\u548c\u300cChatCompletion API\u300d\uff0c\u5176\u5be6\u5c31\u662f\u6c7a\u5b9a AI \u56de\u61c9\u6211\u5011\u7684\u98a8\u683c\u3002<\/p>\n<p>\u95b1\u8b80\u5b8c\u9019\u7bc7\u6587\u7ae0\u5e0c\u671b\u8b80\u8005\u80fd\u5c0d\u9019\u5169\u500b\u6a21\u578b\u4ee5\u53ca\u500b\u5225\u7684API\u56de\u8986\u65b9\u5f0f\u80fd\u6709\u66f4\u591a\u7684\u8a8d\u8b58\uff0c\u4e0d\u540c\u7684\u6a21\u578b\u548cAPI\u9069\u7528\u65bc\u4e0d\u540c\u7684\u5834\u666f\u3002<\/p>\n<p>\u7406\u89e3\u5b83\u5011\u7684\u7279\u9ede\u548c\u9069\u7528\u60c5\u5883\uff0c\u7576\u672a\u4f86\u8b80\u8005\u8981\u61c9\u7528Azure OpenAI\u5728\u81ea\u5df1\u7684\u7a0b\u5f0f\u4e0a\u6642\uff0c\u80fd\u5920\u66f4\u6709\u4fe1\u5fc3\u5730\u505a\u51fa\u6700\u4f73\u9078\u64c7\uff0c\u70ba\u81ea\u5df1\u7684\u61c9\u7528\u7a0b\u5f0f\u767c\u63ee AI \u6700\u5927\u7684\u6f5b\u80fd\u3002<\/p>\n<p><font color=\"#000000\"><br \/><\/font><\/p>\n<p><strong><font color=\"#000000\"><br \/><\/font><\/strong><\/p>\n<p><strong><font color=\"#000000\"><br \/><\/font><\/strong><\/p>\n<p><strong><font color=\"#000000\"><br \/><\/font><\/strong><\/p>\n<p><strong><font color=\"#000000\"><br \/><\/font><\/strong><\/p>\n<p><strong><font color=\"#000000\"><br \/><\/font><\/strong><\/p>\n<p><strong><font color=\"#000000\"><br \/><\/font><\/strong><\/p>\n<p><strong><font color=\"#000000\"><br \/><\/font><\/strong><\/p>\n<p><strong><font color=\"#000000\"><br \/><\/font><\/strong><\/p>\n<p><strong><font color=\"#000000\"><br \/><\/font><\/strong><\/p>\n<p><strong><font color=\"#000000\"><br \/><\/font><\/strong><\/p>\n<p><strong><font color=\"#000000\"><br \/><\/font><\/strong><\/p>\n<p><strong><font color=\"#000000\"><br \/><\/font><\/strong><\/p>\n<p><strong><font color=\"#000000\"><br \/><\/font><\/strong><\/p>\n<p><strong><font color=\"#000000\"><br \/><\/font><\/strong><\/p>\n<p><strong><font color=\"#000000\"><br \/><\/font><\/strong><\/p>\n<p><strong><font color=\"#000000\"><br \/><\/font><\/strong><\/p>\n<p><strong><font color=\"#000000\"><br \/><\/font><\/strong><\/p>\n<p><strong><font color=\"#000000\"><br \/><\/font><\/strong><\/p>\n<p><strong><font color=\"#000000\"><br \/><\/font><\/strong><\/p>\n<p><strong><font color=\"#000000\"><br \/><\/font><\/strong><\/p>\n<p><strong><font color=\"#000000\"><br \/><\/font><\/strong><\/p>\n","protected":false},"excerpt":{"rendered":"<p>\u5fae\u8edf\u4e5d\u6708\u516c\u5e03\u7684 Azure OpenAI \u6700\u65b0\u6d88\u606f\u4e2d\u63d0\u5230: Azure OpenAI\u670d\u52d9\u73fe\u5728\u652f\u63f4 GPT-3.5 Turbo \u6307\u793a\u6a21\u578b\u3002 \u770b\u5230\u9019\u500b\u6d88\u606f\uff0c\u4e0d\u77e5\u9053\u5927&#46;&#46;&#46;<\/p>\n","protected":false},"author":4,"featured_media":1925,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":[],"categories":[22],"tags":[],"_links":{"self":[{"href":"http:\/\/i-services.info\/cspclm\/index.php?rest_route=\/wp\/v2\/posts\/1932"}],"collection":[{"href":"http:\/\/i-services.info\/cspclm\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/i-services.info\/cspclm\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/i-services.info\/cspclm\/index.php?rest_route=\/wp\/v2\/users\/4"}],"replies":[{"embeddable":true,"href":"http:\/\/i-services.info\/cspclm\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=1932"}],"version-history":[{"count":3,"href":"http:\/\/i-services.info\/cspclm\/index.php?rest_route=\/wp\/v2\/posts\/1932\/revisions"}],"predecessor-version":[{"id":1935,"href":"http:\/\/i-services.info\/cspclm\/index.php?rest_route=\/wp\/v2\/posts\/1932\/revisions\/1935"}],"wp:featuredmedia":[{"embeddable":true,"href":"http:\/\/i-services.info\/cspclm\/index.php?rest_route=\/wp\/v2\/media\/1925"}],"wp:attachment":[{"href":"http:\/\/i-services.info\/cspclm\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=1932"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/i-services.info\/cspclm\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=1932"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/i-services.info\/cspclm\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=1932"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}