{"id":63200,"date":"2026-01-21T14:49:50","date_gmt":"2026-01-21T06:49:50","guid":{"rendered":"https:\/\/www.wsisp.com\/helps\/63200.html"},"modified":"2026-01-21T14:49:50","modified_gmt":"2026-01-21T06:49:50","slug":"tensorflow%e6%b7%b1%e5%ba%a6%e5%ad%a6%e4%b9%a0%e6%a1%86%e6%9e%b6%e5%85%a5%e9%97%a8%e6%b5%85%e6%9e%90","status":"publish","type":"post","link":"https:\/\/www.wsisp.com\/helps\/63200.html","title":{"rendered":"TensorFlow\u6df1\u5ea6\u5b66\u4e60\u6846\u67b6\u5165\u95e8\u6d45\u6790"},"content":{"rendered":"<p><img decoding=\"async\" src=\"https:\/\/www.wsisp.com\/helps\/wp-content\/uploads\/2026\/01\/20260121064948-6970770ce5319.png\" alt=\"#\" \/><\/p>\n<h3>\u5f15\u8a00<\/h3>\n<p>TensorFlow\u4f5c\u4e3aGoogle\u5f00\u6e90\u7684\u6df1\u5ea6\u5b66\u4e60\u6846\u67b6&#xff0c;\u51ed\u501f\u5176\u5f3a\u5927\u7684\u5206\u5e03\u5f0f\u8ba1\u7b97\u80fd\u529b\u3001\u4e30\u5bcc\u7684\u5de5\u5177\u751f\u6001\u548c\u5e7f\u6cdb\u7684\u884c\u4e1a\u5e94\u7528&#xff0c;\u6210\u4e3a\u4e86\u5168\u7403\u6700\u6d41\u884c\u7684\u6df1\u5ea6\u5b66\u4e60\u6846\u67b6\u4e4b\u4e00\u3002\u672c\u6587\u5c06\u4eceTensorFlow\u7684\u6838\u5fc3\u6982\u5ff5\u51fa\u53d1&#xff0c;\u7cfb\u7edf\u4ecb\u7ecd\u5176\u57fa\u7840\u7528\u6cd5\u548c\u5b9e\u8df5\u6848\u4f8b&#xff0c;\u5e2e\u52a9\u8bfb\u8005\u5feb\u901f\u638c\u63e1\u8fd9\u4e00\u5f3a\u5927\u7684\u6df1\u5ea6\u5b66\u4e60\u5de5\u5177\u3002<\/p>\n<h3>\u4e00\u3001TensorFlow\u7684\u53d1\u5c55\u5386\u7a0b\u4e0e\u6838\u5fc3\u4f18\u52bf<\/h3>\n<h4>1. \u53d1\u5c55\u5386\u7a0b<\/h4>\n<p>TensorFlow\u7684\u53d1\u5c55\u53ef\u4ee5\u5206\u4e3a\u51e0\u4e2a\u91cd\u8981\u9636\u6bb5&#xff1a;<\/p>\n<ul>\n<li>2015\u5e7411\u6708&#xff1a;Google\u5f00\u6e90TensorFlow 0.1&#xff0c;\u57fa\u4e8e\u4e4b\u524d\u7684DistBelief\u7cfb\u7edf\u5f00\u53d1<\/li>\n<li>2017\u5e742\u6708&#xff1a;TensorFlow 1.0\u53d1\u5e03&#xff0c;\u5f15\u5165\u9759\u6001\u8ba1\u7b97\u56fe\u548cEstimator API<\/li>\n<li>2019\u5e7410\u6708&#xff1a;TensorFlow 2.0\u53d1\u5e03&#xff0c;\u91c7\u7528\u52a8\u6001\u8ba1\u7b97\u56fe&#xff08;Eager Execution&#xff09;\u4f5c\u4e3a\u9ed8\u8ba4\u6a21\u5f0f<\/li>\n<li>2021\u5e745\u6708&#xff1a;TensorFlow 2.5\u53d1\u5e03&#xff0c;\u589e\u5f3a\u4e86\u5bf9GPU\u548cTPU\u7684\u652f\u6301<\/li>\n<li>2023\u5e7410\u6708&#xff1a;TensorFlow 2.14\u53d1\u5e03&#xff0c;\u8fdb\u4e00\u6b65\u4f18\u5316\u4e86\u6027\u80fd\u548c\u6613\u7528\u6027<\/li>\n<\/ul>\n<h4>2. \u6838\u5fc3\u4f18\u52bf<\/h4>\n<h5>2.1 \u5f3a\u5927\u7684\u5206\u5e03\u5f0f\u8ba1\u7b97\u80fd\u529b<\/h5>\n<p>TensorFlow\u539f\u751f\u652f\u6301\u5206\u5e03\u5f0f\u8ba1\u7b97&#xff0c;\u53ef\u4ee5\u8f7b\u677e\u6269\u5c55\u5230\u591a\u53f0\u673a\u5668\u548c\u591a\u4e2a\u8bbe\u5907&#xff08;CPU\u3001GPU\u3001TPU&#xff09;&#xff1a;<\/p>\n<ul>\n<li>\u6570\u636e\u5e76\u884c&#xff1a;\u5c06\u6570\u636e\u5206\u5272\u5230\u4e0d\u540c\u8bbe\u5907\u4e0a\u5e76\u884c\u5904\u7406<\/li>\n<li>\u6a21\u578b\u5e76\u884c&#xff1a;\u5c06\u5927\u578b\u6a21\u578b\u5206\u5272\u5230\u4e0d\u540c\u8bbe\u5907\u4e0a\u6267\u884c<\/li>\n<li>\u6df7\u5408\u5e76\u884c&#xff1a;\u7ed3\u5408\u6570\u636e\u5e76\u884c\u548c\u6a21\u578b\u5e76\u884c\u7684\u4f18\u52bf<\/li>\n<\/ul>\n<h5>2.2 \u5b8c\u6574\u7684\u5de5\u5177\u751f\u6001\u7cfb\u7edf<\/h5>\n<p>TensorFlow\u62e5\u6709\u4e30\u5bcc\u7684\u5de5\u5177\u548c\u5e93&#xff0c;\u652f\u6301\u4ece\u6570\u636e\u9884\u5904\u7406\u5230\u6a21\u578b\u90e8\u7f72\u7684\u5168\u6d41\u7a0b&#xff1a;<\/p>\n<ul>\n<li>TensorFlow Data&#xff1a;\u9ad8\u6548\u7684\u6570\u636e\u52a0\u8f7d\u548c\u9884\u5904\u7406<\/li>\n<li>TensorFlow Hub&#xff1a;\u9884\u8bad\u7ec3\u6a21\u578b\u5e93<\/li>\n<li>TensorFlow Lite&#xff1a;\u79fb\u52a8\u7aef\u548c\u8fb9\u7f18\u8bbe\u5907\u90e8\u7f72<\/li>\n<li>TensorFlow Serving&#xff1a;\u751f\u4ea7\u73af\u5883\u6a21\u578b\u90e8\u7f72<\/li>\n<li>TensorBoard&#xff1a;\u53ef\u89c6\u5316\u8bad\u7ec3\u8fc7\u7a0b\u548c\u6a21\u578b\u6027\u80fd<\/li>\n<\/ul>\n<h5>2.3 \u7075\u6d3b\u7684\u8ba1\u7b97\u56fe\u6267\u884c\u6a21\u5f0f<\/h5>\n<p>TensorFlow 2.0\u4ee5\u540e&#xff0c;\u540c\u65f6\u652f\u6301\u4e24\u79cd\u8ba1\u7b97\u56fe\u6267\u884c\u6a21\u5f0f&#xff1a;<\/p>\n<ul>\n<li>\u52a8\u6001\u8ba1\u7b97\u56fe&#xff08;Eager Execution&#xff09;&#xff1a;\u5373\u65f6\u6267\u884c&#xff0c;\u4fbf\u4e8e\u8c03\u8bd5<\/li>\n<li>\u9759\u6001\u8ba1\u7b97\u56fe&#xff08;Graph Execution&#xff09;&#xff1a;\u4f18\u5316\u6267\u884c&#xff0c;\u4fbf\u4e8e\u90e8\u7f72<\/li>\n<\/ul>\n<h5>2.4 \u5e7f\u6cdb\u7684\u884c\u4e1a\u5e94\u7528<\/h5>\n<p>TensorFlow\u5728\u5404\u4e2a\u884c\u4e1a\u90fd\u6709\u5e7f\u6cdb\u7684\u5e94\u7528&#xff1a;<\/p>\n<ul>\n<li>\u8ba1\u7b97\u673a\u89c6\u89c9&#xff1a;\u56fe\u50cf\u5206\u7c7b\u3001\u76ee\u6807\u68c0\u6d4b\u3001\u56fe\u50cf\u5206\u5272<\/li>\n<li>\u81ea\u7136\u8bed\u8a00\u5904\u7406&#xff1a;\u673a\u5668\u7ffb\u8bd1\u3001\u6587\u672c\u751f\u6210\u3001\u60c5\u611f\u5206\u6790<\/li>\n<li>\u63a8\u8350\u7cfb\u7edf&#xff1a;\u4e2a\u6027\u5316\u63a8\u8350\u3001\u70b9\u51fb\u7387\u9884\u6d4b<\/li>\n<li>\u5f3a\u5316\u5b66\u4e60&#xff1a;\u6e38\u620fAI\u3001\u673a\u5668\u4eba\u63a7\u5236<\/li>\n<\/ul>\n<h3>\u4e8c\u3001TensorFlow\u6838\u5fc3\u6982\u5ff5<\/h3>\n<h4>1. \u5f20\u91cf&#xff08;Tensor&#xff09;<\/h4>\n<p>\u5f20\u91cf\u662fTensorFlow\u7684\u57fa\u672c\u6570\u636e\u7ed3\u6784&#xff0c;\u7c7b\u4f3c\u4e8e\u591a\u7ef4\u6570\u7ec4&#xff1a;<\/p>\n<p><span class=\"token keyword\">import<\/span> tensorflow <span class=\"token keyword\">as<\/span> tf<br \/>\n<span class=\"token keyword\">import<\/span> numpy <span class=\"token keyword\">as<\/span> np<\/p>\n<p><span class=\"token comment\"># \u521b\u5efa\u5f20\u91cf<\/span><br \/>\nscalar <span class=\"token operator\">&#061;<\/span> tf<span class=\"token punctuation\">.<\/span>constant<span class=\"token punctuation\">(<\/span><span class=\"token number\">3.14<\/span><span class=\"token punctuation\">)<\/span>  <span class=\"token comment\"># \u6807\u91cf&#xff08;0\u7ef4\u5f20\u91cf&#xff09;<\/span><br \/>\nvector <span class=\"token operator\">&#061;<\/span> tf<span class=\"token punctuation\">.<\/span>constant<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">[<\/span><span class=\"token number\">1<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">2<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">3<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">)<\/span>  <span class=\"token comment\"># \u5411\u91cf&#xff08;1\u7ef4\u5f20\u91cf&#xff09;<\/span><br \/>\nmatrix <span class=\"token operator\">&#061;<\/span> tf<span class=\"token punctuation\">.<\/span>constant<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">[<\/span><span class=\"token number\">1<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">2<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token punctuation\">[<\/span><span class=\"token number\">3<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">4<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">)<\/span>  <span class=\"token comment\"># \u77e9\u9635&#xff08;2\u7ef4\u5f20\u91cf&#xff09;<\/span><br \/>\ntensor_3d <span class=\"token operator\">&#061;<\/span> tf<span class=\"token punctuation\">.<\/span>constant<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">[<\/span><span class=\"token number\">1<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">2<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token punctuation\">[<\/span><span class=\"token number\">3<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">4<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">[<\/span><span class=\"token number\">5<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">6<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token punctuation\">[<\/span><span class=\"token number\">7<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">8<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">)<\/span>  <span class=\"token comment\"># 3\u7ef4\u5f20\u91cf<\/span><\/p>\n<p><span class=\"token comment\"># \u83b7\u53d6\u5f20\u91cf\u7684\u5f62\u72b6\u3001\u7c7b\u578b\u548c\u8bbe\u5907<\/span><br \/>\n<span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u5f20\u91cf\u5f62\u72b6: <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>tensor_3d<span class=\"token punctuation\">.<\/span>shape<span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><br \/>\n<span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u5f20\u91cf\u7c7b\u578b: <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>tensor_3d<span class=\"token punctuation\">.<\/span>dtype<span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><br \/>\n<span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u5f20\u91cf\u8bbe\u5907: <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>tensor_3d<span class=\"token punctuation\">.<\/span>device<span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># NumPy\u6570\u7ec4\u4e0e\u5f20\u91cf\u7684\u8f6c\u6362<\/span><br \/>\nnp_array <span class=\"token operator\">&#061;<\/span> np<span class=\"token punctuation\">.<\/span>array<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">[<\/span><span class=\"token number\">1<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">2<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">3<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">)<\/span><br \/>\ntf_tensor <span class=\"token operator\">&#061;<\/span> tf<span class=\"token punctuation\">.<\/span>convert_to_tensor<span class=\"token punctuation\">(<\/span>np_array<span class=\"token punctuation\">)<\/span><br \/>\nback_to_np <span class=\"token operator\">&#061;<\/span> tf_tensor<span class=\"token punctuation\">.<\/span>numpy<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># GPU\u652f\u6301<\/span><br \/>\n<span class=\"token keyword\">if<\/span> tf<span class=\"token punctuation\">.<\/span>config<span class=\"token punctuation\">.<\/span>list_physical_devices<span class=\"token punctuation\">(<\/span><span class=\"token string\">&#039;GPU&#039;<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n    <span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string\">&#034;GPU\u53ef\u7528&#034;<\/span><span class=\"token punctuation\">)<\/span><br \/>\n    <span class=\"token keyword\">with<\/span> tf<span class=\"token punctuation\">.<\/span>device<span class=\"token punctuation\">(<\/span><span class=\"token string\">&#039;GPU:0&#039;<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n        gpu_tensor <span class=\"token operator\">&#061;<\/span> tf<span class=\"token punctuation\">.<\/span>constant<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">[<\/span><span class=\"token number\">1<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">2<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">3<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">)<\/span><br \/>\n        <span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;GPU\u5f20\u91cf\u8bbe\u5907: <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>gpu_tensor<span class=\"token punctuation\">.<\/span>device<span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><br \/>\n<span class=\"token keyword\">else<\/span><span class=\"token punctuation\">:<\/span><br \/>\n    <span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string\">&#034;GPU\u4e0d\u53ef\u7528&#034;<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<h4>2. \u81ea\u52a8\u5fae\u5206&#xff08;tf.GradientTape&#xff09;<\/h4>\n<p>TensorFlow\u4f7f\u7528tf.GradientTape\u8bb0\u5f55\u64cd\u4f5c&#xff0c;\u81ea\u52a8\u8ba1\u7b97\u68af\u5ea6&#xff1a;<\/p>\n<p><span class=\"token keyword\">import<\/span> tensorflow <span class=\"token keyword\">as<\/span> tf<\/p>\n<p><span class=\"token comment\"># \u521b\u5efa\u9700\u8981\u8ba1\u7b97\u68af\u5ea6\u7684\u5f20\u91cf<\/span><br \/>\na <span class=\"token operator\">&#061;<\/span> tf<span class=\"token punctuation\">.<\/span>Variable<span class=\"token punctuation\">(<\/span><span class=\"token number\">2.0<\/span><span class=\"token punctuation\">)<\/span><br \/>\nb <span class=\"token operator\">&#061;<\/span> tf<span class=\"token punctuation\">.<\/span>Variable<span class=\"token punctuation\">(<\/span><span class=\"token number\">3.0<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u4f7f\u7528GradientTape\u8bb0\u5f55\u8ba1\u7b97\u8fc7\u7a0b<\/span><br \/>\n<span class=\"token keyword\">with<\/span> tf<span class=\"token punctuation\">.<\/span>GradientTape<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span> <span class=\"token keyword\">as<\/span> tape<span class=\"token punctuation\">:<\/span><br \/>\n    c <span class=\"token operator\">&#061;<\/span> a <span class=\"token operator\">*<\/span> b <span class=\"token operator\">&#043;<\/span> tf<span class=\"token punctuation\">.<\/span>square<span class=\"token punctuation\">(<\/span>a<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u8ba1\u7b97\u68af\u5ea6<\/span><br \/>\ngradients <span class=\"token operator\">&#061;<\/span> tape<span class=\"token punctuation\">.<\/span>gradient<span class=\"token punctuation\">(<\/span>c<span class=\"token punctuation\">,<\/span> <span class=\"token punctuation\">[<\/span>a<span class=\"token punctuation\">,<\/span> b<span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u6253\u5370\u68af\u5ea6<\/span><br \/>\n<span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;dc\/da: <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>gradients<span class=\"token punctuation\">[<\/span><span class=\"token number\">0<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span>  <span class=\"token comment\"># \u8f93\u51fa: dc\/da: tf.Tensor(8.0, shape&#061;(), dtype&#061;float32)<\/span><br \/>\n<span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;dc\/db: <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>gradients<span class=\"token punctuation\">[<\/span><span class=\"token number\">1<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span>  <span class=\"token comment\"># \u8f93\u51fa: dc\/db: tf.Tensor(2.0, shape&#061;(), dtype&#061;float32)<\/span><\/p>\n<h4>3. \u795e\u7ecf\u7f51\u7edc\u6784\u5efa&#xff08;tf.keras&#xff09;<\/h4>\n<p>TensorFlow 2.0\u4ee5\u540e&#xff0c;\u63a8\u8350\u4f7f\u7528tf.keras API\u6784\u5efa\u795e\u7ecf\u7f51\u7edc&#xff1a;<\/p>\n<p><span class=\"token keyword\">import<\/span> tensorflow <span class=\"token keyword\">as<\/span> tf<br \/>\n<span class=\"token keyword\">from<\/span> tensorflow<span class=\"token punctuation\">.<\/span>keras <span class=\"token keyword\">import<\/span> layers<span class=\"token punctuation\">,<\/span> models<\/p>\n<p><span class=\"token comment\"># \u4f7f\u7528Sequential API\u6784\u5efa\u7b80\u5355\u6a21\u578b<\/span><br \/>\nmodel <span class=\"token operator\">&#061;<\/span> models<span class=\"token punctuation\">.<\/span>Sequential<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">[<\/span><br \/>\n    layers<span class=\"token punctuation\">.<\/span>Dense<span class=\"token punctuation\">(<\/span><span class=\"token number\">64<\/span><span class=\"token punctuation\">,<\/span> activation<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#039;relu&#039;<\/span><span class=\"token punctuation\">,<\/span> input_shape<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">(<\/span><span class=\"token number\">784<\/span><span class=\"token punctuation\">,<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">,<\/span>  <span class=\"token comment\"># \u8f93\u5165\u5c42\u5230\u9690\u85cf\u5c42<\/span><br \/>\n    layers<span class=\"token punctuation\">.<\/span>Dense<span class=\"token punctuation\">(<\/span><span class=\"token number\">32<\/span><span class=\"token punctuation\">,<\/span> activation<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#039;relu&#039;<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">,<\/span>  <span class=\"token comment\"># \u9690\u85cf\u5c42<\/span><br \/>\n    layers<span class=\"token punctuation\">.<\/span>Dense<span class=\"token punctuation\">(<\/span><span class=\"token number\">10<\/span><span class=\"token punctuation\">,<\/span> activation<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#039;softmax&#039;<\/span><span class=\"token punctuation\">)<\/span>  <span class=\"token comment\"># \u8f93\u51fa\u5c42<\/span><br \/>\n<span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u4f7f\u7528Functional API\u6784\u5efa\u590d\u6742\u6a21\u578b<\/span><br \/>\ninputs <span class=\"token operator\">&#061;<\/span> tf<span class=\"token punctuation\">.<\/span>keras<span class=\"token punctuation\">.<\/span>Input<span class=\"token punctuation\">(<\/span>shape<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">(<\/span><span class=\"token number\">784<\/span><span class=\"token punctuation\">,<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">)<\/span><br \/>\nhidden1 <span class=\"token operator\">&#061;<\/span> layers<span class=\"token punctuation\">.<\/span>Dense<span class=\"token punctuation\">(<\/span><span class=\"token number\">64<\/span><span class=\"token punctuation\">,<\/span> activation<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#039;relu&#039;<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">(<\/span>inputs<span class=\"token punctuation\">)<\/span><br \/>\nhidden2 <span class=\"token operator\">&#061;<\/span> layers<span class=\"token punctuation\">.<\/span>Dense<span class=\"token punctuation\">(<\/span><span class=\"token number\">32<\/span><span class=\"token punctuation\">,<\/span> activation<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#039;relu&#039;<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">(<\/span>hidden1<span class=\"token punctuation\">)<\/span><br \/>\noutputs <span class=\"token operator\">&#061;<\/span> layers<span class=\"token punctuation\">.<\/span>Dense<span class=\"token punctuation\">(<\/span><span class=\"token number\">10<\/span><span class=\"token punctuation\">,<\/span> activation<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#039;softmax&#039;<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">(<\/span>hidden2<span class=\"token punctuation\">)<\/span><\/p>\n<p>complex_model <span class=\"token operator\">&#061;<\/span> models<span class=\"token punctuation\">.<\/span>Model<span class=\"token punctuation\">(<\/span>inputs<span class=\"token operator\">&#061;<\/span>inputs<span class=\"token punctuation\">,<\/span> outputs<span class=\"token operator\">&#061;<\/span>outputs<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u67e5\u770b\u6a21\u578b\u7ed3\u6784<\/span><br \/>\nmodel<span class=\"token punctuation\">.<\/span>summary<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><br \/>\ncomplex_model<span class=\"token punctuation\">.<\/span>summary<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<h4>4. \u635f\u5931\u51fd\u6570\u4e0e\u4f18\u5316\u5668<\/h4>\n<p>TensorFlow\u63d0\u4f9b\u4e86\u4e30\u5bcc\u7684\u635f\u5931\u51fd\u6570\u548c\u4f18\u5316\u5668&#xff1a;<\/p>\n<p><span class=\"token keyword\">import<\/span> tensorflow <span class=\"token keyword\">as<\/span> tf<\/p>\n<p><span class=\"token comment\"># \u5b9a\u4e49\u635f\u5931\u51fd\u6570<\/span><br \/>\nloss_fn <span class=\"token operator\">&#061;<\/span> tf<span class=\"token punctuation\">.<\/span>keras<span class=\"token punctuation\">.<\/span>losses<span class=\"token punctuation\">.<\/span>SparseCategoricalCrossentropy<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span>  <span class=\"token comment\"># \u9002\u7528\u4e8e\u6574\u6570\u6807\u7b7e<\/span><br \/>\n<span class=\"token comment\"># \u6216\u4f7f\u7528\u5206\u7c7b\u4ea4\u53c9\u71b5&#xff08;\u9002\u7528\u4e8e\u72ec\u70ed\u7f16\u7801\u6807\u7b7e&#xff09;<\/span><br \/>\n<span class=\"token comment\"># loss_fn &#061; tf.keras.losses.CategoricalCrossentropy()<\/span><\/p>\n<p><span class=\"token comment\"># \u5b9a\u4e49\u4f18\u5316\u5668<\/span><br \/>\noptimizer <span class=\"token operator\">&#061;<\/span> tf<span class=\"token punctuation\">.<\/span>keras<span class=\"token punctuation\">.<\/span>optimizers<span class=\"token punctuation\">.<\/span>Adam<span class=\"token punctuation\">(<\/span>learning_rate<span class=\"token operator\">&#061;<\/span><span class=\"token number\">0.001<\/span><span class=\"token punctuation\">)<\/span><br \/>\n<span class=\"token comment\"># \u6216\u4f7f\u7528SGD\u4f18\u5316\u5668<\/span><br \/>\n<span class=\"token comment\"># optimizer &#061; tf.keras.optimizers.SGD(learning_rate&#061;0.01, momentum&#061;0.9)<\/span><\/p>\n<p><span class=\"token comment\"># \u5b9a\u4e49\u8bc4\u4f30\u6307\u6807<\/span><br \/>\nmetrics <span class=\"token operator\">&#061;<\/span> <span class=\"token punctuation\">[<\/span>tf<span class=\"token punctuation\">.<\/span>keras<span class=\"token punctuation\">.<\/span>metrics<span class=\"token punctuation\">.<\/span>SparseCategoricalAccuracy<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">]<\/span><\/p>\n<h3>\u4e09\u3001TensorFlow\u5b9e\u8df5\u6848\u4f8b&#xff1a;\u56fe\u50cf\u5206\u7c7b<\/h3>\n<h4>1. \u6570\u636e\u96c6\u51c6\u5907<\/h4>\n<p>\u4f7f\u7528TensorFlow Datasets\u52a0\u8f7dMNIST\u624b\u5199\u6570\u5b57\u6570\u636e\u96c6&#xff1a;<\/p>\n<p><span class=\"token keyword\">import<\/span> tensorflow <span class=\"token keyword\">as<\/span> tf<br \/>\n<span class=\"token keyword\">import<\/span> tensorflow_datasets <span class=\"token keyword\">as<\/span> tfds<\/p>\n<p><span class=\"token comment\"># \u52a0\u8f7dMNIST\u6570\u636e\u96c6<\/span><br \/>\ndataset<span class=\"token punctuation\">,<\/span> info <span class=\"token operator\">&#061;<\/span> tfds<span class=\"token punctuation\">.<\/span>load<span class=\"token punctuation\">(<\/span><span class=\"token string\">&#039;mnist&#039;<\/span><span class=\"token punctuation\">,<\/span> with_info<span class=\"token operator\">&#061;<\/span><span class=\"token boolean\">True<\/span><span class=\"token punctuation\">,<\/span> as_supervised<span class=\"token operator\">&#061;<\/span><span class=\"token boolean\">True<\/span><span class=\"token punctuation\">)<\/span><br \/>\ntrain_dataset<span class=\"token punctuation\">,<\/span> test_dataset <span class=\"token operator\">&#061;<\/span> dataset<span class=\"token punctuation\">[<\/span><span class=\"token string\">&#039;train&#039;<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">,<\/span> dataset<span class=\"token punctuation\">[<\/span><span class=\"token string\">&#039;test&#039;<\/span><span class=\"token punctuation\">]<\/span><\/p>\n<p><span class=\"token comment\"># \u83b7\u53d6\u6570\u636e\u96c6\u4fe1\u606f<\/span><br \/>\n<span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u8bad\u7ec3\u96c6\u5927\u5c0f: <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>info<span class=\"token punctuation\">.<\/span>splits<span class=\"token punctuation\">[<\/span><span class=\"token string\">&#039;train&#039;<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">.<\/span>num_examples<span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><br \/>\n<span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u6d4b\u8bd5\u96c6\u5927\u5c0f: <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>info<span class=\"token punctuation\">.<\/span>splits<span class=\"token punctuation\">[<\/span><span class=\"token string\">&#039;test&#039;<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">.<\/span>num_examples<span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><br \/>\n<span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u56fe\u50cf\u5f62\u72b6: <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>info<span class=\"token punctuation\">.<\/span>features<span class=\"token punctuation\">[<\/span><span class=\"token string\">&#039;image&#039;<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">.<\/span>shape<span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><br \/>\n<span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u6807\u7b7e\u6570\u91cf: <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>info<span class=\"token punctuation\">.<\/span>features<span class=\"token punctuation\">[<\/span><span class=\"token string\">&#039;label&#039;<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">.<\/span>num_classes<span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u6570\u636e\u9884\u5904\u7406<\/span><br \/>\n<span class=\"token keyword\">def<\/span> <span class=\"token function\">preprocess<\/span><span class=\"token punctuation\">(<\/span>image<span class=\"token punctuation\">,<\/span> label<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n    <span class=\"token comment\"># \u5c06\u56fe\u50cf\u8f6c\u6362\u4e3afloat32\u7c7b\u578b<\/span><br \/>\n    image <span class=\"token operator\">&#061;<\/span> tf<span class=\"token punctuation\">.<\/span>cast<span class=\"token punctuation\">(<\/span>image<span class=\"token punctuation\">,<\/span> tf<span class=\"token punctuation\">.<\/span>float32<span class=\"token punctuation\">)<\/span><br \/>\n    <span class=\"token comment\"># \u5f52\u4e00\u5316\u5230[-1, 1]\u8303\u56f4<\/span><br \/>\n    image <span class=\"token operator\">&#061;<\/span> <span class=\"token punctuation\">(<\/span>image <span class=\"token operator\">\/<\/span> <span class=\"token number\">127.5<\/span><span class=\"token punctuation\">)<\/span> <span class=\"token operator\">&#8211;<\/span> <span class=\"token number\">1<\/span><br \/>\n    <span class=\"token comment\"># \u5c55\u5e73\u56fe\u50cf<\/span><br \/>\n    image <span class=\"token operator\">&#061;<\/span> tf<span class=\"token punctuation\">.<\/span>reshape<span class=\"token punctuation\">(<\/span>image<span class=\"token punctuation\">,<\/span> <span class=\"token punctuation\">(<\/span><span class=\"token number\">784<\/span><span class=\"token punctuation\">,<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">)<\/span><br \/>\n    <span class=\"token keyword\">return<\/span> image<span class=\"token punctuation\">,<\/span> label<\/p>\n<p><span class=\"token comment\"># \u5e94\u7528\u9884\u5904\u7406\u5e76\u521b\u5efa\u6279\u6b21<\/span><br \/>\nbatch_size <span class=\"token operator\">&#061;<\/span> <span class=\"token number\">64<\/span><br \/>\ntrain_dataset <span class=\"token operator\">&#061;<\/span> train_dataset<span class=\"token punctuation\">.<\/span><span class=\"token builtin\">map<\/span><span class=\"token punctuation\">(<\/span>preprocess<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">.<\/span>shuffle<span class=\"token punctuation\">(<\/span><span class=\"token number\">10000<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">.<\/span>batch<span class=\"token punctuation\">(<\/span>batch_size<span class=\"token punctuation\">)<\/span><br \/>\ntest_dataset <span class=\"token operator\">&#061;<\/span> test_dataset<span class=\"token punctuation\">.<\/span><span class=\"token builtin\">map<\/span><span class=\"token punctuation\">(<\/span>preprocess<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">.<\/span>batch<span class=\"token punctuation\">(<\/span>batch_size<span class=\"token punctuation\">)<\/span><\/p>\n<h4>2. \u6784\u5efa\u795e\u7ecf\u7f51\u7edc\u6a21\u578b<\/h4>\n<p><span class=\"token keyword\">import<\/span> tensorflow <span class=\"token keyword\">as<\/span> tf<br \/>\n<span class=\"token keyword\">from<\/span> tensorflow<span class=\"token punctuation\">.<\/span>keras <span class=\"token keyword\">import<\/span> layers<span class=\"token punctuation\">,<\/span> models<\/p>\n<p><span class=\"token comment\"># \u6784\u5efa\u795e\u7ecf\u7f51\u7edc\u6a21\u578b<\/span><br \/>\nmodel <span class=\"token operator\">&#061;<\/span> models<span class=\"token punctuation\">.<\/span>Sequential<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">[<\/span><br \/>\n    layers<span class=\"token punctuation\">.<\/span>Dense<span class=\"token punctuation\">(<\/span><span class=\"token number\">256<\/span><span class=\"token punctuation\">,<\/span> activation<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#039;relu&#039;<\/span><span class=\"token punctuation\">,<\/span> input_shape<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">(<\/span><span class=\"token number\">784<\/span><span class=\"token punctuation\">,<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">,<\/span><br \/>\n                kernel_regularizer<span class=\"token operator\">&#061;<\/span>tf<span class=\"token punctuation\">.<\/span>keras<span class=\"token punctuation\">.<\/span>regularizers<span class=\"token punctuation\">.<\/span>l2<span class=\"token punctuation\">(<\/span><span class=\"token number\">0.001<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">,<\/span>  <span class=\"token comment\"># \u5e26L2\u6b63\u5219\u5316\u7684\u9690\u85cf\u5c42<\/span><br \/>\n    layers<span class=\"token punctuation\">.<\/span>Dropout<span class=\"token punctuation\">(<\/span><span class=\"token number\">0.3<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">,<\/span>  <span class=\"token comment\"># Dropout\u5c42\u9632\u6b62\u8fc7\u62df\u5408<\/span><br \/>\n    layers<span class=\"token punctuation\">.<\/span>Dense<span class=\"token punctuation\">(<\/span><span class=\"token number\">128<\/span><span class=\"token punctuation\">,<\/span> activation<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#039;relu&#039;<\/span><span class=\"token punctuation\">,<\/span><br \/>\n                kernel_regularizer<span class=\"token operator\">&#061;<\/span>tf<span class=\"token punctuation\">.<\/span>keras<span class=\"token punctuation\">.<\/span>regularizers<span class=\"token punctuation\">.<\/span>l2<span class=\"token punctuation\">(<\/span><span class=\"token number\">0.001<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">,<\/span>  <span class=\"token comment\"># \u5e26L2\u6b63\u5219\u5316\u7684\u9690\u85cf\u5c42<\/span><br \/>\n    layers<span class=\"token punctuation\">.<\/span>Dropout<span class=\"token punctuation\">(<\/span><span class=\"token number\">0.3<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">,<\/span>  <span class=\"token comment\"># Dropout\u5c42\u9632\u6b62\u8fc7\u62df\u5408<\/span><br \/>\n    layers<span class=\"token punctuation\">.<\/span>Dense<span class=\"token punctuation\">(<\/span><span class=\"token number\">10<\/span><span class=\"token punctuation\">,<\/span> activation<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#039;softmax&#039;<\/span><span class=\"token punctuation\">)<\/span>  <span class=\"token comment\"># \u8f93\u51fa\u5c42&#xff0c;10\u4e2a\u7c7b\u522b<\/span><br \/>\n<span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u7f16\u8bd1\u6a21\u578b<\/span><br \/>\nmodel<span class=\"token punctuation\">.<\/span><span class=\"token builtin\">compile<\/span><span class=\"token punctuation\">(<\/span>optimizer<span class=\"token operator\">&#061;<\/span>tf<span class=\"token punctuation\">.<\/span>keras<span class=\"token punctuation\">.<\/span>optimizers<span class=\"token punctuation\">.<\/span>Adam<span class=\"token punctuation\">(<\/span>learning_rate<span class=\"token operator\">&#061;<\/span><span class=\"token number\">0.001<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">,<\/span><br \/>\n             loss<span class=\"token operator\">&#061;<\/span>tf<span class=\"token punctuation\">.<\/span>keras<span class=\"token punctuation\">.<\/span>losses<span class=\"token punctuation\">.<\/span>SparseCategoricalCrossentropy<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">,<\/span><br \/>\n             metrics<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">[<\/span>tf<span class=\"token punctuation\">.<\/span>keras<span class=\"token punctuation\">.<\/span>metrics<span class=\"token punctuation\">.<\/span>SparseCategoricalAccuracy<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<h4>3. \u8bad\u7ec3\u6a21\u578b<\/h4>\n<p><span class=\"token comment\"># \u5b9a\u4e49\u56de\u8c03\u51fd\u6570<\/span><br \/>\ncallbacks <span class=\"token operator\">&#061;<\/span> <span class=\"token punctuation\">[<\/span><br \/>\n    tf<span class=\"token punctuation\">.<\/span>keras<span class=\"token punctuation\">.<\/span>callbacks<span class=\"token punctuation\">.<\/span>EarlyStopping<span class=\"token punctuation\">(<\/span>patience<span class=\"token operator\">&#061;<\/span><span class=\"token number\">3<\/span><span class=\"token punctuation\">,<\/span> monitor<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#039;val_loss&#039;<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">,<\/span>  <span class=\"token comment\"># \u65e9\u505c\u6cd5\u9632\u6b62\u8fc7\u62df\u5408<\/span><br \/>\n    tf<span class=\"token punctuation\">.<\/span>keras<span class=\"token punctuation\">.<\/span>callbacks<span class=\"token punctuation\">.<\/span>TensorBoard<span class=\"token punctuation\">(<\/span>log_dir<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#039;.\/logs&#039;<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">,<\/span>  <span class=\"token comment\"># TensorBoard\u53ef\u89c6\u5316<\/span><br \/>\n    tf<span class=\"token punctuation\">.<\/span>keras<span class=\"token punctuation\">.<\/span>callbacks<span class=\"token punctuation\">.<\/span>ModelCheckpoint<span class=\"token punctuation\">(<\/span>filepath<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#039;best_model.h5&#039;<\/span><span class=\"token punctuation\">,<\/span><br \/>\n                                      monitor<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#039;val_sparse_categorical_accuracy&#039;<\/span><span class=\"token punctuation\">,<\/span><br \/>\n                                      save_best_only<span class=\"token operator\">&#061;<\/span><span class=\"token boolean\">True<\/span><span class=\"token punctuation\">)<\/span>  <span class=\"token comment\"># \u4fdd\u5b58\u6700\u4f73\u6a21\u578b<\/span><br \/>\n<span class=\"token punctuation\">]<\/span><\/p>\n<p><span class=\"token comment\"># \u8bad\u7ec3\u6a21\u578b<\/span><br \/>\nhistory <span class=\"token operator\">&#061;<\/span> model<span class=\"token punctuation\">.<\/span>fit<span class=\"token punctuation\">(<\/span>train_dataset<span class=\"token punctuation\">,<\/span><br \/>\n                   epochs<span class=\"token operator\">&#061;<\/span><span class=\"token number\">20<\/span><span class=\"token punctuation\">,<\/span><br \/>\n                   validation_data<span class=\"token operator\">&#061;<\/span>test_dataset<span class=\"token punctuation\">,<\/span><br \/>\n                   callbacks<span class=\"token operator\">&#061;<\/span>callbacks<span class=\"token punctuation\">)<\/span><\/p>\n<h4>4. \u8bc4\u4f30\u6a21\u578b<\/h4>\n<p><span class=\"token comment\"># \u52a0\u8f7d\u6700\u4f73\u6a21\u578b<\/span><br \/>\nbest_model <span class=\"token operator\">&#061;<\/span> tf<span class=\"token punctuation\">.<\/span>keras<span class=\"token punctuation\">.<\/span>models<span class=\"token punctuation\">.<\/span>load_model<span class=\"token punctuation\">(<\/span><span class=\"token string\">&#039;best_model.h5&#039;<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u8bc4\u4f30\u6a21\u578b<\/span><br \/>\nloss<span class=\"token punctuation\">,<\/span> accuracy <span class=\"token operator\">&#061;<\/span> best_model<span class=\"token punctuation\">.<\/span>evaluate<span class=\"token punctuation\">(<\/span>test_dataset<span class=\"token punctuation\">)<\/span><br \/>\n<span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u6d4b\u8bd5\u635f\u5931: <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>loss<span class=\"token punctuation\">:<\/span><span class=\"token format-spec\">.4f<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><br \/>\n<span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u6d4b\u8bd5\u51c6\u786e\u7387: <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>accuracy<span class=\"token punctuation\">:<\/span><span class=\"token format-spec\">.4f<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u8fdb\u884c\u9884\u6d4b<\/span><br \/>\ntest_images<span class=\"token punctuation\">,<\/span> test_labels <span class=\"token operator\">&#061;<\/span> <span class=\"token builtin\">next<\/span><span class=\"token punctuation\">(<\/span><span class=\"token builtin\">iter<\/span><span class=\"token punctuation\">(<\/span>test_dataset<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">)<\/span><br \/>\npredictions <span class=\"token operator\">&#061;<\/span> best_model<span class=\"token punctuation\">.<\/span>predict<span class=\"token punctuation\">(<\/span>test_images<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u67e5\u770b\u9884\u6d4b\u7ed3\u679c<\/span><br \/>\n<span class=\"token keyword\">import<\/span> numpy <span class=\"token keyword\">as<\/span> np<br \/>\nfirst_image <span class=\"token operator\">&#061;<\/span> test_images<span class=\"token punctuation\">[<\/span><span class=\"token number\">0<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">.<\/span>numpy<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">.<\/span>reshape<span class=\"token punctuation\">(<\/span><span class=\"token number\">28<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">28<\/span><span class=\"token punctuation\">)<\/span><br \/>\nfirst_label <span class=\"token operator\">&#061;<\/span> test_labels<span class=\"token punctuation\">[<\/span><span class=\"token number\">0<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">.<\/span>numpy<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><br \/>\nfirst_prediction <span class=\"token operator\">&#061;<\/span> np<span class=\"token punctuation\">.<\/span>argmax<span class=\"token punctuation\">(<\/span>predictions<span class=\"token punctuation\">[<\/span><span class=\"token number\">0<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u771f\u5b9e\u6807\u7b7e: <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>first_label<span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><br \/>\n<span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u9884\u6d4b\u6807\u7b7e: <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>first_prediction<span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><br \/>\n<span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u9884\u6d4b\u6982\u7387: <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>predictions<span class=\"token punctuation\">[<\/span><span class=\"token number\">0<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">[<\/span>first_prediction<span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">:<\/span><span class=\"token format-spec\">.4f<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><\/p>\n<h3>\u56db\u3001TensorFlow\u9ad8\u7ea7\u7279\u6027<\/h3>\n<h4>1. \u81ea\u5b9a\u4e49\u8bad\u7ec3\u5faa\u73af<\/h4>\n<p>\u5bf9\u4e8e\u66f4\u590d\u6742\u7684\u8bad\u7ec3\u9700\u6c42&#xff0c;\u53ef\u4ee5\u4f7f\u7528\u81ea\u5b9a\u4e49\u8bad\u7ec3\u5faa\u73af&#xff1a;<\/p>\n<p><span class=\"token keyword\">import<\/span> tensorflow <span class=\"token keyword\">as<\/span> tf<\/p>\n<p><span class=\"token comment\"># \u5b9a\u4e49\u635f\u5931\u51fd\u6570\u548c\u4f18\u5316\u5668<\/span><br \/>\nloss_fn <span class=\"token operator\">&#061;<\/span> tf<span class=\"token punctuation\">.<\/span>keras<span class=\"token punctuation\">.<\/span>losses<span class=\"token punctuation\">.<\/span>SparseCategoricalCrossentropy<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><br \/>\noptimizer <span class=\"token operator\">&#061;<\/span> tf<span class=\"token punctuation\">.<\/span>keras<span class=\"token punctuation\">.<\/span>optimizers<span class=\"token punctuation\">.<\/span>Adam<span class=\"token punctuation\">(<\/span>learning_rate<span class=\"token operator\">&#061;<\/span><span class=\"token number\">0.001<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u5b9a\u4e49\u8bc4\u4f30\u6307\u6807<\/span><br \/>\ntrain_loss <span class=\"token operator\">&#061;<\/span> tf<span class=\"token punctuation\">.<\/span>keras<span class=\"token punctuation\">.<\/span>metrics<span class=\"token punctuation\">.<\/span>Mean<span class=\"token punctuation\">(<\/span>name<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#039;train_loss&#039;<\/span><span class=\"token punctuation\">)<\/span><br \/>\ntrain_accuracy <span class=\"token operator\">&#061;<\/span> tf<span class=\"token punctuation\">.<\/span>keras<span class=\"token punctuation\">.<\/span>metrics<span class=\"token punctuation\">.<\/span>SparseCategoricalAccuracy<span class=\"token punctuation\">(<\/span>name<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#039;train_accuracy&#039;<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>test_loss <span class=\"token operator\">&#061;<\/span> tf<span class=\"token punctuation\">.<\/span>keras<span class=\"token punctuation\">.<\/span>metrics<span class=\"token punctuation\">.<\/span>Mean<span class=\"token punctuation\">(<\/span>name<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#039;test_loss&#039;<\/span><span class=\"token punctuation\">)<\/span><br \/>\ntest_accuracy <span class=\"token operator\">&#061;<\/span> tf<span class=\"token punctuation\">.<\/span>keras<span class=\"token punctuation\">.<\/span>metrics<span class=\"token punctuation\">.<\/span>SparseCategoricalAccuracy<span class=\"token punctuation\">(<\/span>name<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#039;test_accuracy&#039;<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u5b9a\u4e49\u8bad\u7ec3\u6b65\u9aa4<\/span><br \/>\n<span class=\"token decorator annotation punctuation\">&#064;tf<span class=\"token punctuation\">.<\/span>function<\/span>  <span class=\"token comment\"># \u8f6c\u6362\u4e3a\u9759\u6001\u8ba1\u7b97\u56fe&#xff0c;\u63d0\u9ad8\u6027\u80fd<\/span><br \/>\n<span class=\"token keyword\">def<\/span> <span class=\"token function\">train_step<\/span><span class=\"token punctuation\">(<\/span>images<span class=\"token punctuation\">,<\/span> labels<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n    <span class=\"token keyword\">with<\/span> tf<span class=\"token punctuation\">.<\/span>GradientTape<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span> <span class=\"token keyword\">as<\/span> tape<span class=\"token punctuation\">:<\/span><br \/>\n        <span class=\"token comment\"># \u524d\u5411\u4f20\u64ad<\/span><br \/>\n        predictions <span class=\"token operator\">&#061;<\/span> model<span class=\"token punctuation\">(<\/span>images<span class=\"token punctuation\">,<\/span> training<span class=\"token operator\">&#061;<\/span><span class=\"token boolean\">True<\/span><span class=\"token punctuation\">)<\/span><br \/>\n        <span class=\"token comment\"># \u8ba1\u7b97\u635f\u5931<\/span><br \/>\n        loss <span class=\"token operator\">&#061;<\/span> loss_fn<span class=\"token punctuation\">(<\/span>labels<span class=\"token punctuation\">,<\/span> predictions<span class=\"token punctuation\">)<\/span><\/p>\n<p>    <span class=\"token comment\"># \u8ba1\u7b97\u68af\u5ea6<\/span><br \/>\n    gradients <span class=\"token operator\">&#061;<\/span> tape<span class=\"token punctuation\">.<\/span>gradient<span class=\"token punctuation\">(<\/span>loss<span class=\"token punctuation\">,<\/span> model<span class=\"token punctuation\">.<\/span>trainable_variables<span class=\"token punctuation\">)<\/span><br \/>\n    <span class=\"token comment\"># \u66f4\u65b0\u53c2\u6570<\/span><br \/>\n    optimizer<span class=\"token punctuation\">.<\/span>apply_gradients<span class=\"token punctuation\">(<\/span><span class=\"token builtin\">zip<\/span><span class=\"token punctuation\">(<\/span>gradients<span class=\"token punctuation\">,<\/span> model<span class=\"token punctuation\">.<\/span>trainable_variables<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>    <span class=\"token comment\"># \u66f4\u65b0\u6307\u6807<\/span><br \/>\n    train_loss<span class=\"token punctuation\">(<\/span>loss<span class=\"token punctuation\">)<\/span><br \/>\n    train_accuracy<span class=\"token punctuation\">(<\/span>labels<span class=\"token punctuation\">,<\/span> predictions<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u5b9a\u4e49\u6d4b\u8bd5\u6b65\u9aa4<\/span><br \/>\n<span class=\"token decorator annotation punctuation\">&#064;tf<span class=\"token punctuation\">.<\/span>function<\/span><br \/>\n<span class=\"token keyword\">def<\/span> <span class=\"token function\">test_step<\/span><span class=\"token punctuation\">(<\/span>images<span class=\"token punctuation\">,<\/span> labels<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n    <span class=\"token comment\"># \u524d\u5411\u4f20\u64ad<\/span><br \/>\n    predictions <span class=\"token operator\">&#061;<\/span> model<span class=\"token punctuation\">(<\/span>images<span class=\"token punctuation\">,<\/span> training<span class=\"token operator\">&#061;<\/span><span class=\"token boolean\">False<\/span><span class=\"token punctuation\">)<\/span><br \/>\n    <span class=\"token comment\"># \u8ba1\u7b97\u635f\u5931<\/span><br \/>\n    t_loss <span class=\"token operator\">&#061;<\/span> loss_fn<span class=\"token punctuation\">(<\/span>labels<span class=\"token punctuation\">,<\/span> predictions<span class=\"token punctuation\">)<\/span><\/p>\n<p>    <span class=\"token comment\"># \u66f4\u65b0\u6307\u6807<\/span><br \/>\n    test_loss<span class=\"token punctuation\">(<\/span>t_loss<span class=\"token punctuation\">)<\/span><br \/>\n    test_accuracy<span class=\"token punctuation\">(<\/span>labels<span class=\"token punctuation\">,<\/span> predictions<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u8bad\u7ec3\u5faa\u73af<\/span><br \/>\nepochs <span class=\"token operator\">&#061;<\/span> <span class=\"token number\">10<\/span><\/p>\n<p><span class=\"token keyword\">for<\/span> epoch <span class=\"token keyword\">in<\/span> <span class=\"token builtin\">range<\/span><span class=\"token punctuation\">(<\/span>epochs<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n    <span class=\"token comment\"># \u91cd\u7f6e\u6307\u6807<\/span><br \/>\n    train_loss<span class=\"token punctuation\">.<\/span>reset_states<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><br \/>\n    train_accuracy<span class=\"token punctuation\">.<\/span>reset_states<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><br \/>\n    test_loss<span class=\"token punctuation\">.<\/span>reset_states<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><br \/>\n    test_accuracy<span class=\"token punctuation\">.<\/span>reset_states<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>    <span class=\"token comment\"># \u8bad\u7ec3<\/span><br \/>\n    <span class=\"token keyword\">for<\/span> images<span class=\"token punctuation\">,<\/span> labels <span class=\"token keyword\">in<\/span> train_dataset<span class=\"token punctuation\">:<\/span><br \/>\n        train_step<span class=\"token punctuation\">(<\/span>images<span class=\"token punctuation\">,<\/span> labels<span class=\"token punctuation\">)<\/span><\/p>\n<p>    <span class=\"token comment\"># \u6d4b\u8bd5<\/span><br \/>\n    <span class=\"token keyword\">for<\/span> images<span class=\"token punctuation\">,<\/span> labels <span class=\"token keyword\">in<\/span> test_dataset<span class=\"token punctuation\">:<\/span><br \/>\n        test_step<span class=\"token punctuation\">(<\/span>images<span class=\"token punctuation\">,<\/span> labels<span class=\"token punctuation\">)<\/span><\/p>\n<p>    <span class=\"token comment\"># \u6253\u5370\u7ed3\u679c<\/span><br \/>\n    <span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;Epoch <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>epoch <span class=\"token operator\">&#043;<\/span> <span class=\"token number\">1<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">, &#034;<\/span><\/span><br \/>\n          <span class=\"token string-interpolation\"><span class=\"token string\">f&#034;Loss: <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>train_loss<span class=\"token punctuation\">.<\/span>result<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><span class=\"token format-spec\">.4f<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">, &#034;<\/span><\/span><br \/>\n          <span class=\"token string-interpolation\"><span class=\"token string\">f&#034;Accuracy: <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>train_accuracy<span class=\"token punctuation\">.<\/span>result<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><span class=\"token format-spec\">.4f<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">, &#034;<\/span><\/span><br \/>\n          <span class=\"token string-interpolation\"><span class=\"token string\">f&#034;Test Loss: <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>test_loss<span class=\"token punctuation\">.<\/span>result<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><span class=\"token format-spec\">.4f<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">, &#034;<\/span><\/span><br \/>\n          <span class=\"token string-interpolation\"><span class=\"token string\">f&#034;Test Accuracy: <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>test_accuracy<span class=\"token punctuation\">.<\/span>result<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><span class=\"token format-spec\">.4f<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><\/p>\n<h4>2. \u5206\u5e03\u5f0f\u8bad\u7ec3<\/h4>\n<p>TensorFlow\u652f\u6301\u591a\u79cd\u5206\u5e03\u5f0f\u8bad\u7ec3\u7b56\u7565&#xff1a;<\/p>\n<p><span class=\"token keyword\">import<\/span> tensorflow <span class=\"token keyword\">as<\/span> tf<\/p>\n<p><span class=\"token comment\"># \u5355\u673a\u5668\u591aGPU\u8bad\u7ec3<\/span><br \/>\nstrategy <span class=\"token operator\">&#061;<\/span> tf<span class=\"token punctuation\">.<\/span>distribute<span class=\"token punctuation\">.<\/span>MirroredStrategy<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token keyword\">with<\/span> strategy<span class=\"token punctuation\">.<\/span>scope<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n    <span class=\"token comment\"># \u5728\u7b56\u7565\u8303\u56f4\u5185\u6784\u5efa\u6a21\u578b<\/span><br \/>\n    model <span class=\"token operator\">&#061;<\/span> tf<span class=\"token punctuation\">.<\/span>keras<span class=\"token punctuation\">.<\/span>Sequential<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">[<\/span><br \/>\n        tf<span class=\"token punctuation\">.<\/span>keras<span class=\"token punctuation\">.<\/span>layers<span class=\"token punctuation\">.<\/span>Dense<span class=\"token punctuation\">(<\/span><span class=\"token number\">128<\/span><span class=\"token punctuation\">,<\/span> activation<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#039;relu&#039;<\/span><span class=\"token punctuation\">,<\/span> input_shape<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">(<\/span><span class=\"token number\">784<\/span><span class=\"token punctuation\">,<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">,<\/span><br \/>\n        tf<span class=\"token punctuation\">.<\/span>keras<span class=\"token punctuation\">.<\/span>layers<span class=\"token punctuation\">.<\/span>Dense<span class=\"token punctuation\">(<\/span><span class=\"token number\">10<\/span><span class=\"token punctuation\">,<\/span> activation<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#039;softmax&#039;<\/span><span class=\"token punctuation\">)<\/span><br \/>\n    <span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>    <span class=\"token comment\"># \u7f16\u8bd1\u6a21\u578b<\/span><br \/>\n    model<span class=\"token punctuation\">.<\/span><span class=\"token builtin\">compile<\/span><span class=\"token punctuation\">(<\/span>optimizer<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#039;adam&#039;<\/span><span class=\"token punctuation\">,<\/span><br \/>\n                 loss<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#039;sparse_categorical_crossentropy&#039;<\/span><span class=\"token punctuation\">,<\/span><br \/>\n                 metrics<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">[<\/span><span class=\"token string\">&#039;accuracy&#039;<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u8bad\u7ec3\u6a21\u578b<\/span><br \/>\nmodel<span class=\"token punctuation\">.<\/span>fit<span class=\"token punctuation\">(<\/span>train_dataset<span class=\"token punctuation\">,<\/span> epochs<span class=\"token operator\">&#061;<\/span><span class=\"token number\">10<\/span><span class=\"token punctuation\">,<\/span> validation_data<span class=\"token operator\">&#061;<\/span>test_dataset<span class=\"token punctuation\">)<\/span><\/p>\n<h4>3. \u6a21\u578b\u90e8\u7f72<\/h4>\n<p>TensorFlow\u63d0\u4f9b\u4e86\u591a\u79cd\u6a21\u578b\u90e8\u7f72\u65b9\u5f0f&#xff1a;<\/p>\n<p><span class=\"token keyword\">import<\/span> tensorflow <span class=\"token keyword\">as<\/span> tf<\/p>\n<p><span class=\"token comment\"># \u4fdd\u5b58\u4e3aSavedModel\u683c\u5f0f<\/span><br \/>\ntf<span class=\"token punctuation\">.<\/span>saved_model<span class=\"token punctuation\">.<\/span>save<span class=\"token punctuation\">(<\/span>model<span class=\"token punctuation\">,<\/span> <span class=\"token string\">&#039;saved_model&#039;<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u52a0\u8f7dSavedModel<\/span><br \/>\nloaded_model <span class=\"token operator\">&#061;<\/span> tf<span class=\"token punctuation\">.<\/span>saved_model<span class=\"token punctuation\">.<\/span>load<span class=\"token punctuation\">(<\/span><span class=\"token string\">&#039;saved_model&#039;<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u8f6c\u6362\u4e3aTensorFlow Lite\u683c\u5f0f&#xff08;\u7528\u4e8e\u79fb\u52a8\u7aef\u548c\u8fb9\u7f18\u8bbe\u5907&#xff09;<\/span><br \/>\nconverter <span class=\"token operator\">&#061;<\/span> tf<span class=\"token punctuation\">.<\/span>lite<span class=\"token punctuation\">.<\/span>TFLiteConverter<span class=\"token punctuation\">.<\/span>from_saved_model<span class=\"token punctuation\">(<\/span><span class=\"token string\">&#039;saved_model&#039;<\/span><span class=\"token punctuation\">)<\/span><br \/>\ntflite_model <span class=\"token operator\">&#061;<\/span> converter<span class=\"token punctuation\">.<\/span>convert<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u4fdd\u5b58TensorFlow Lite\u6a21\u578b<\/span><br \/>\n<span class=\"token keyword\">with<\/span> <span class=\"token builtin\">open<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string\">&#039;model.tflite&#039;<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token string\">&#039;wb&#039;<\/span><span class=\"token punctuation\">)<\/span> <span class=\"token keyword\">as<\/span> f<span class=\"token punctuation\">:<\/span><br \/>\n    f<span class=\"token punctuation\">.<\/span>write<span class=\"token punctuation\">(<\/span>tflite_model<span class=\"token punctuation\">)<\/span><\/p>\n<h3>\u4e94\u3001\u5b66\u4e60\u8d44\u6e90\u548c\u6700\u4f73\u5b9e\u8df5<\/h3>\n<h4>1. \u5b98\u65b9\u8d44\u6e90<\/h4>\n<ul>\n<li>TensorFlow\u5b98\u65b9\u6587\u6863<\/li>\n<li>TensorFlow\u6559\u7a0b<\/li>\n<li>TensorFlow Hub &#8211; \u9884\u8bad\u7ec3\u6a21\u578b\u5e93<\/li>\n<li>TensorBoard\u6559\u7a0b<\/li>\n<\/ul>\n<h4>2. \u5b66\u4e60\u8def\u5f84<\/h4>\n<li>\u638c\u63e1\u57fa\u7840\u5f20\u91cf\u64cd\u4f5c\u548c\u81ea\u52a8\u5fae\u5206<\/li>\n<li>\u5b66\u4e60\u4f7f\u7528tf.keras\u6784\u5efa\u795e\u7ecf\u7f51\u7edc<\/li>\n<li>\u719f\u6089\u5e38\u7528\u7684\u635f\u5931\u51fd\u6570\u3001\u4f18\u5316\u5668\u548c\u8bc4\u4f30\u6307\u6807<\/li>\n<li>\u5b9e\u8df5\u56fe\u50cf\u5206\u7c7b\u3001\u6587\u672c\u5206\u7c7b\u7b49\u7ecf\u5178\u4efb\u52a1<\/li>\n<li>\u5b66\u4e60\u9ad8\u7ea7\u7279\u6027&#xff08;\u81ea\u5b9a\u4e49\u8bad\u7ec3\u5faa\u73af\u3001\u5206\u5e03\u5f0f\u8bad\u7ec3\u7b49&#xff09;<\/li>\n<li>\u638c\u63e1\u6a21\u578b\u90e8\u7f72\u548c\u751f\u4ea7\u73af\u5883\u5e94\u7528<\/li>\n<h4>3. \u6700\u4f73\u5b9e\u8df5<\/h4>\n<ul>\n<li>\u4f7f\u7528tf.keras API&#xff1a;TensorFlow 2.0\u4ee5\u540e&#xff0c;tf.keras\u662f\u63a8\u8350\u7684API<\/li>\n<li>\u6570\u636e\u9884\u5904\u7406&#xff1a;\u4f7f\u7528tf.data API\u8fdb\u884c\u9ad8\u6548\u7684\u6570\u636e\u52a0\u8f7d\u548c\u9884\u5904\u7406<\/li>\n<li>\u6b63\u5219\u5316&#xff1a;\u5408\u7406\u4f7f\u7528Dropout\u3001L1\/L2\u6b63\u5219\u5316\u9632\u6b62\u8fc7\u62df\u5408<\/li>\n<li>\u5b66\u4e60\u7387\u8c03\u5ea6&#xff1a;\u4f7f\u7528\u5b66\u4e60\u7387\u8870\u51cf\u7b56\u7565\u63d0\u9ad8\u6a21\u578b\u6027\u80fd<\/li>\n<li>\u65e9\u505c\u6cd5&#xff1a;\u76d1\u63a7\u9a8c\u8bc1\u635f\u5931&#xff0c;\u907f\u514d\u8fc7\u62df\u5408<\/li>\n<li>TensorBoard\u53ef\u89c6\u5316&#xff1a;\u5b9e\u65f6\u76d1\u63a7\u8bad\u7ec3\u8fc7\u7a0b<\/li>\n<\/ul>\n<p><span class=\"token comment\"># \u4f7f\u7528\u5b66\u4e60\u7387\u8c03\u5ea6\u5668<\/span><br \/>\nlr_scheduler <span class=\"token operator\">&#061;<\/span> tf<span class=\"token punctuation\">.<\/span>keras<span class=\"token punctuation\">.<\/span>callbacks<span class=\"token punctuation\">.<\/span>ReduceLROnPlateau<span class=\"token punctuation\">(<\/span>monitor<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#039;val_loss&#039;<\/span><span class=\"token punctuation\">,<\/span><br \/>\n                                                    factor<span class=\"token operator\">&#061;<\/span><span class=\"token number\">0.1<\/span><span class=\"token punctuation\">,<\/span><br \/>\n                                                    patience<span class=\"token operator\">&#061;<\/span><span class=\"token number\">3<\/span><span class=\"token punctuation\">,<\/span><br \/>\n                                                    min_lr<span class=\"token operator\">&#061;<\/span><span class=\"token number\">0.00001<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u4f7f\u7528\u65e9\u505c\u6cd5<\/span><br \/>\nearly_stopping <span class=\"token operator\">&#061;<\/span> tf<span class=\"token punctuation\">.<\/span>keras<span class=\"token punctuation\">.<\/span>callbacks<span class=\"token punctuation\">.<\/span>EarlyStopping<span class=\"token punctuation\">(<\/span>monitor<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#039;val_loss&#039;<\/span><span class=\"token punctuation\">,<\/span><br \/>\n                                                  patience<span class=\"token operator\">&#061;<\/span><span class=\"token number\">5<\/span><span class=\"token punctuation\">,<\/span><br \/>\n                                                  restore_best_weights<span class=\"token operator\">&#061;<\/span><span class=\"token boolean\">True<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u8bad\u7ec3\u6a21\u578b<\/span><br \/>\nmodel<span class=\"token punctuation\">.<\/span>fit<span class=\"token punctuation\">(<\/span>train_dataset<span class=\"token punctuation\">,<\/span><br \/>\n         epochs<span class=\"token operator\">&#061;<\/span><span class=\"token number\">50<\/span><span class=\"token punctuation\">,<\/span><br \/>\n         validation_data<span class=\"token operator\">&#061;<\/span>test_dataset<span class=\"token punctuation\">,<\/span><br \/>\n         callbacks<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">[<\/span>lr_scheduler<span class=\"token punctuation\">,<\/span> early_stopping<span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<h3>\u516d\u3001TensorFlow\u4e0ePyTorch\u7684\u6bd4\u8f83<\/h3>\n<table>\n<tr>\u7279\u6027TensorFlowPyTorch<\/tr>\n<tbody>\n<tr>\n<td>\u5f00\u53d1\u516c\u53f8<\/td>\n<td>Google<\/td>\n<td>Facebook\/Meta<\/td>\n<\/tr>\n<tr>\n<td>\u9ed8\u8ba4\u6267\u884c\u6a21\u5f0f<\/td>\n<td>\u52a8\u6001\u8ba1\u7b97\u56fe&#xff08;Eager Execution&#xff09;<\/td>\n<td>\u52a8\u6001\u8ba1\u7b97\u56fe<\/td>\n<\/tr>\n<tr>\n<td>\u9759\u6001\u8ba1\u7b97\u56fe\u652f\u6301<\/td>\n<td>\u652f\u6301&#xff08;\u901a\u8fc7tf.function&#xff09;<\/td>\n<td>\u4e0d\u76f4\u63a5\u652f\u6301<\/td>\n<\/tr>\n<tr>\n<td>\u5206\u5e03\u5f0f\u8ba1\u7b97<\/td>\n<td>\u5f3a\u5927&#xff0c;\u652f\u6301\u591a\u79cd\u7b56\u7565<\/td>\n<td>\u652f\u6301&#xff0c;\u4f46\u76f8\u5bf9\u7b80\u5355<\/td>\n<\/tr>\n<tr>\n<td>\u6a21\u578b\u90e8\u7f72<\/td>\n<td>\u5b8c\u5584&#xff08;TensorFlow Serving\u3001TFLite&#xff09;<\/td>\n<td>\u76f8\u5bf9\u8f83\u5c11<\/td>\n<\/tr>\n<tr>\n<td>\u5de5\u5177\u751f\u6001<\/td>\n<td>\u4e30\u5bcc&#xff08;TF Hub\u3001TF Data\u7b49&#xff09;<\/td>\n<td>\u6b63\u5728\u53d1\u5c55<\/td>\n<\/tr>\n<tr>\n<td>\u793e\u533a\u652f\u6301<\/td>\n<td>\u5e7f\u6cdb<\/td>\n<td>\u6d3b\u8dc3&#xff0c;\u7279\u522b\u662f\u5b66\u672f\u754c<\/td>\n<\/tr>\n<tr>\n<td>\u5b66\u4e60\u66f2\u7ebf<\/td>\n<td>\u8f83\u9661\u5ced<\/td>\n<td>\u8f83\u5e73\u7f13<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<h3>\u4e03\u3001\u603b\u7ed3<\/h3>\n<p>TensorFlow\u4f5c\u4e3a\u4e00\u6b3e\u6210\u719f\u7684\u6df1\u5ea6\u5b66\u4e60\u6846\u67b6&#xff0c;\u51ed\u501f\u5176\u5f3a\u5927\u7684\u529f\u80fd\u548c\u4e30\u5bcc\u7684\u751f\u6001&#xff0c;\u5728\u5de5\u4e1a\u754c\u548c\u5b66\u672f\u754c\u90fd\u6709\u7740\u5e7f\u6cdb\u7684\u5e94\u7528\u3002\u672c\u6587\u4eceTensorFlow\u7684\u6838\u5fc3\u6982\u5ff5\u51fa\u53d1&#xff0c;\u7cfb\u7edf\u4ecb\u7ecd\u4e86\u5176\u57fa\u7840\u7528\u6cd5\u548c\u5b9e\u8df5\u6848\u4f8b&#xff0c;\u5e0c\u671b\u80fd\u5e2e\u52a9\u8bfb\u8005\u5feb\u901f\u5165\u95e8TensorFlow\u3002<\/p>\n<p>\u968f\u7740\u6df1\u5ea6\u5b66\u4e60\u6280\u672f\u7684\u4e0d\u65ad\u53d1\u5c55&#xff0c;TensorFlow\u4e5f\u5728\u6301\u7eed\u6f14\u8fdb&#xff0c;\u63a8\u51fa\u4e86\u66f4\u591a\u9ad8\u7ea7\u7279\u6027\u548c\u4f18\u5316\u3002\u5efa\u8bae\u8bfb\u8005\u5728\u638c\u63e1\u57fa\u7840\u540e&#xff0c;\u8fdb\u4e00\u6b65\u5b66\u4e60TensorFlow\u7684\u9ad8\u7ea7\u529f\u80fd&#xff0c;\u5982\u5206\u5e03\u5f0f\u8bad\u7ec3\u3001\u6a21\u578b\u91cf\u5316\u548c\u751f\u4ea7\u73af\u5883\u90e8\u7f72&#xff0c;\u4ee5\u5e94\u5bf9\u66f4\u590d\u6742\u7684\u6df1\u5ea6\u5b66\u4e60\u4efb\u52a1\u3002<\/p>\n<p>\u6700\u540e&#xff0c;TensorFlow\u7684\u5b66\u4e60\u662f\u4e00\u4e2a\u5b9e\u8df5\u7684\u8fc7\u7a0b&#xff0c;\u5efa\u8bae\u8bfb\u8005\u901a\u8fc7\u52a8\u624b\u5b9e\u8df5\u6765\u52a0\u6df1\u7406\u89e3&#xff0c;\u9010\u6b65\u638c\u63e1\u8fd9\u4e00\u5f3a\u5927\u7684\u6df1\u5ea6\u5b66\u4e60\u6846\u67b6\u3002<\/p>\n","protected":false},"excerpt":{"rendered":"<p>\u5f15\u8a00<br \/>\nTensorFlow\u4f5c\u4e3aGoogle\u5f00\u6e90\u7684\u6df1\u5ea6\u5b66\u4e60\u6846\u67b6&#xff0c;\u51ed\u501f\u5176\u5f3a\u5927\u7684\u5206\u5e03\u5f0f\u8ba1\u7b97\u80fd\u529b\u3001\u4e30\u5bcc\u7684\u5de5\u5177\u751f\u6001\u548c\u5e7f\u6cdb\u7684\u884c\u4e1a\u5e94\u7528&#xff0c;\u6210\u4e3a\u4e86\u5168\u7403\u6700\u6d41\u884c\u7684\u6df1\u5ea6\u5b66\u4e60\u6846\u67b6\u4e4b\u4e00\u3002\u672c\u6587\u5c06\u4eceTensorFlow\u7684\u6838\u5fc3\u6982\u5ff5\u51fa\u53d1&#xff0c;\u7cfb\u7edf\u4ecb\u7ecd\u5176\u57fa\u7840\u7528\u6cd5\u548c\u5b9e\u8df5\u6848\u4f8b&#xff0c;\u5e2e\u52a9\u8bfb\u8005\u5feb\u901f\u638c\u63e1\u8fd9\u4e00\u5f3a\u5927\u7684\u6df1\u5ea6\u5b66\u4e60\u5de5\u5177\u3002<br \/>\n\u4e00\u3001TensorFlow\u7684\u53d1\u5c55\u5386\u7a0b\u4e0e\u6838\u5fc3\u4f18\u52bf<br \/>\n1. \u53d1\u5c55\u5386\u7a0b<br \/>\nTensorFlow\u7684\u53d1\u5c55\u53ef\u4ee5\u5206\u4e3a\u51e0\u4e2a\u91cd\u8981\u9636\u6bb5&#xff1a;<br \/>\n2015\u5e7411<\/p>\n","protected":false},"author":2,"featured_media":63199,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[587,6686,6684,6685,2756,3677,86],"topic":[],"class_list":["post-63200","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-server","tag-587","tag-6686","tag-6684","tag-6685","tag-neo4j","tag-tensorflow","tag-86"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v20.3 - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>TensorFlow\u6df1\u5ea6\u5b66\u4e60\u6846\u67b6\u5165\u95e8\u6d45\u6790 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.wsisp.com\/helps\/63200.html\" \/>\n<meta property=\"og:locale\" content=\"zh_CN\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"TensorFlow\u6df1\u5ea6\u5b66\u4e60\u6846\u67b6\u5165\u95e8\u6d45\u6790 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\" \/>\n<meta property=\"og:description\" content=\"\u5f15\u8a00 TensorFlow\u4f5c\u4e3aGoogle\u5f00\u6e90\u7684\u6df1\u5ea6\u5b66\u4e60\u6846\u67b6&#xff0c;\u51ed\u501f\u5176\u5f3a\u5927\u7684\u5206\u5e03\u5f0f\u8ba1\u7b97\u80fd\u529b\u3001\u4e30\u5bcc\u7684\u5de5\u5177\u751f\u6001\u548c\u5e7f\u6cdb\u7684\u884c\u4e1a\u5e94\u7528&#xff0c;\u6210\u4e3a\u4e86\u5168\u7403\u6700\u6d41\u884c\u7684\u6df1\u5ea6\u5b66\u4e60\u6846\u67b6\u4e4b\u4e00\u3002\u672c\u6587\u5c06\u4eceTensorFlow\u7684\u6838\u5fc3\u6982\u5ff5\u51fa\u53d1&#xff0c;\u7cfb\u7edf\u4ecb\u7ecd\u5176\u57fa\u7840\u7528\u6cd5\u548c\u5b9e\u8df5\u6848\u4f8b&#xff0c;\u5e2e\u52a9\u8bfb\u8005\u5feb\u901f\u638c\u63e1\u8fd9\u4e00\u5f3a\u5927\u7684\u6df1\u5ea6\u5b66\u4e60\u5de5\u5177\u3002 \u4e00\u3001TensorFlow\u7684\u53d1\u5c55\u5386\u7a0b\u4e0e\u6838\u5fc3\u4f18\u52bf 1. \u53d1\u5c55\u5386\u7a0b TensorFlow\u7684\u53d1\u5c55\u53ef\u4ee5\u5206\u4e3a\u51e0\u4e2a\u91cd\u8981\u9636\u6bb5&#xff1a; 2015\u5e7411\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.wsisp.com\/helps\/63200.html\" \/>\n<meta property=\"og:site_name\" content=\"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\" \/>\n<meta property=\"article:published_time\" content=\"2026-01-21T06:49:50+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/www.wsisp.com\/helps\/wp-content\/uploads\/2026\/01\/20260121064948-6970770ce5319.png\" \/>\n<meta name=\"author\" content=\"admin\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"\u4f5c\u8005\" \/>\n\t<meta name=\"twitter:data1\" content=\"admin\" \/>\n\t<meta name=\"twitter:label2\" content=\"\u9884\u8ba1\u9605\u8bfb\u65f6\u95f4\" \/>\n\t<meta name=\"twitter:data2\" content=\"6 \u5206\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/63200.html\",\"url\":\"https:\/\/www.wsisp.com\/helps\/63200.html\",\"name\":\"TensorFlow\u6df1\u5ea6\u5b66\u4e60\u6846\u67b6\u5165\u95e8\u6d45\u6790 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\",\"isPartOf\":{\"@id\":\"https:\/\/www.wsisp.com\/helps\/#website\"},\"datePublished\":\"2026-01-21T06:49:50+00:00\",\"dateModified\":\"2026-01-21T06:49:50+00:00\",\"author\":{\"@id\":\"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41\"},\"breadcrumb\":{\"@id\":\"https:\/\/www.wsisp.com\/helps\/63200.html#breadcrumb\"},\"inLanguage\":\"zh-Hans\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/www.wsisp.com\/helps\/63200.html\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/63200.html#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"\u9996\u9875\",\"item\":\"https:\/\/www.wsisp.com\/helps\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"TensorFlow\u6df1\u5ea6\u5b66\u4e60\u6846\u67b6\u5165\u95e8\u6d45\u6790\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/#website\",\"url\":\"https:\/\/www.wsisp.com\/helps\/\",\"name\":\"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\",\"description\":\"\u9999\u6e2f\u670d\u52a1\u5668_\u9999\u6e2f\u4e91\u670d\u52a1\u5668\u8d44\u8baf_\u670d\u52a1\u5668\u5e2e\u52a9\u6587\u6863_\u670d\u52a1\u5668\u6559\u7a0b\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/www.wsisp.com\/helps\/?s={search_term_string}\"},\"query-input\":\"required name=search_term_string\"}],\"inLanguage\":\"zh-Hans\"},{\"@type\":\"Person\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41\",\"name\":\"admin\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"zh-Hans\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery\",\"contentUrl\":\"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery\",\"caption\":\"admin\"},\"sameAs\":[\"http:\/\/wp.wsisp.com\"],\"url\":\"https:\/\/www.wsisp.com\/helps\/author\/admin\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"TensorFlow\u6df1\u5ea6\u5b66\u4e60\u6846\u67b6\u5165\u95e8\u6d45\u6790 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.wsisp.com\/helps\/63200.html","og_locale":"zh_CN","og_type":"article","og_title":"TensorFlow\u6df1\u5ea6\u5b66\u4e60\u6846\u67b6\u5165\u95e8\u6d45\u6790 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","og_description":"\u5f15\u8a00 TensorFlow\u4f5c\u4e3aGoogle\u5f00\u6e90\u7684\u6df1\u5ea6\u5b66\u4e60\u6846\u67b6&#xff0c;\u51ed\u501f\u5176\u5f3a\u5927\u7684\u5206\u5e03\u5f0f\u8ba1\u7b97\u80fd\u529b\u3001\u4e30\u5bcc\u7684\u5de5\u5177\u751f\u6001\u548c\u5e7f\u6cdb\u7684\u884c\u4e1a\u5e94\u7528&#xff0c;\u6210\u4e3a\u4e86\u5168\u7403\u6700\u6d41\u884c\u7684\u6df1\u5ea6\u5b66\u4e60\u6846\u67b6\u4e4b\u4e00\u3002\u672c\u6587\u5c06\u4eceTensorFlow\u7684\u6838\u5fc3\u6982\u5ff5\u51fa\u53d1&#xff0c;\u7cfb\u7edf\u4ecb\u7ecd\u5176\u57fa\u7840\u7528\u6cd5\u548c\u5b9e\u8df5\u6848\u4f8b&#xff0c;\u5e2e\u52a9\u8bfb\u8005\u5feb\u901f\u638c\u63e1\u8fd9\u4e00\u5f3a\u5927\u7684\u6df1\u5ea6\u5b66\u4e60\u5de5\u5177\u3002 \u4e00\u3001TensorFlow\u7684\u53d1\u5c55\u5386\u7a0b\u4e0e\u6838\u5fc3\u4f18\u52bf 1. \u53d1\u5c55\u5386\u7a0b TensorFlow\u7684\u53d1\u5c55\u53ef\u4ee5\u5206\u4e3a\u51e0\u4e2a\u91cd\u8981\u9636\u6bb5&#xff1a; 2015\u5e7411","og_url":"https:\/\/www.wsisp.com\/helps\/63200.html","og_site_name":"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","article_published_time":"2026-01-21T06:49:50+00:00","og_image":[{"url":"https:\/\/www.wsisp.com\/helps\/wp-content\/uploads\/2026\/01\/20260121064948-6970770ce5319.png"}],"author":"admin","twitter_card":"summary_large_image","twitter_misc":{"\u4f5c\u8005":"admin","\u9884\u8ba1\u9605\u8bfb\u65f6\u95f4":"6 \u5206"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/www.wsisp.com\/helps\/63200.html","url":"https:\/\/www.wsisp.com\/helps\/63200.html","name":"TensorFlow\u6df1\u5ea6\u5b66\u4e60\u6846\u67b6\u5165\u95e8\u6d45\u6790 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","isPartOf":{"@id":"https:\/\/www.wsisp.com\/helps\/#website"},"datePublished":"2026-01-21T06:49:50+00:00","dateModified":"2026-01-21T06:49:50+00:00","author":{"@id":"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41"},"breadcrumb":{"@id":"https:\/\/www.wsisp.com\/helps\/63200.html#breadcrumb"},"inLanguage":"zh-Hans","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.wsisp.com\/helps\/63200.html"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/www.wsisp.com\/helps\/63200.html#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"\u9996\u9875","item":"https:\/\/www.wsisp.com\/helps"},{"@type":"ListItem","position":2,"name":"TensorFlow\u6df1\u5ea6\u5b66\u4e60\u6846\u67b6\u5165\u95e8\u6d45\u6790"}]},{"@type":"WebSite","@id":"https:\/\/www.wsisp.com\/helps\/#website","url":"https:\/\/www.wsisp.com\/helps\/","name":"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","description":"\u9999\u6e2f\u670d\u52a1\u5668_\u9999\u6e2f\u4e91\u670d\u52a1\u5668\u8d44\u8baf_\u670d\u52a1\u5668\u5e2e\u52a9\u6587\u6863_\u670d\u52a1\u5668\u6559\u7a0b","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.wsisp.com\/helps\/?s={search_term_string}"},"query-input":"required name=search_term_string"}],"inLanguage":"zh-Hans"},{"@type":"Person","@id":"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41","name":"admin","image":{"@type":"ImageObject","inLanguage":"zh-Hans","@id":"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/image\/","url":"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery","contentUrl":"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery","caption":"admin"},"sameAs":["http:\/\/wp.wsisp.com"],"url":"https:\/\/www.wsisp.com\/helps\/author\/admin"}]}},"_links":{"self":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/posts\/63200","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/comments?post=63200"}],"version-history":[{"count":0,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/posts\/63200\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/media\/63199"}],"wp:attachment":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/media?parent=63200"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/categories?post=63200"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/tags?post=63200"},{"taxonomy":"topic","embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/topic?post=63200"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}