{"id":2567,"date":"2026-02-15T18:56:47","date_gmt":"2026-02-15T10:56:47","guid":{"rendered":"https:\/\/ai.ziyuanzz.online\/?page_id=2567"},"modified":"2026-03-28T09:32:47","modified_gmt":"2026-03-28T01:32:47","slug":"keras","status":"publish","type":"page","link":"https:\/\/ai.ziyuanzz.online\/index.php\/keras\/","title":{"rendered":"Keras"},"content":{"rendered":"
\n
\"\"<\/figure>\n<\/div>\n\n\n

Keras<\/h2>\n\n\n\n

Python\u7248\u672c\u7684TensorFlow\u6df1\u5ea6\u5b66\u4e60API<\/p>\n\n\n\n

\n
\u8bbf\u95ee\u5b98\u7f51 ><\/a><\/div>\n<\/div>\n\n\n\n
<\/div>\n\n\n\n
\"\"<\/figure>\n\n\n\n
<\/div>\n\n\n\n

Keras\u662f\u4ec0\u4e48<\/h2>\n\n\n\n

Keras \u662f\u5f00\u6e90\u7684\u6df1\u5ea6\u5b66\u4e60\u6846\u67b6\uff0c\u4e13\u4e3a\u4eba\u7c7b\u8bbe\u8ba1\uff0c\u6ce8\u91cd\u6613\u7528\u6027\u3001\u7075\u6d3b\u6027\u548c\u9ad8\u6548\u6027\u3002Keras \u652f\u6301 TensorFlow<\/a>\u3001JAX<\/a> \u548c PyTorch<\/a> \u7b49\u591a\u79cd\u540e\u7aef\uff0c\u652f\u6301\u5f00\u53d1\u8005\u65e0\u7f1d\u5207\u6362\u6846\u67b6\u3002Keras \u63d0\u4f9b\u7b80\u6d01\u7684 API \u548c\u4e30\u5bcc\u7684\u9884\u8bad\u7ec3\u6a21\u578b\uff0c\u9002\u5408\u4ece\u521d\u5b66\u8005\u5230\u9ad8\u7ea7\u5f00\u53d1\u8005\u7684\u5404\u79cd\u9700\u6c42\u3002Keras \u5177\u5907\u6a21\u5757\u5316\u8bbe\u8ba1\u3001\u9ad8\u6027\u80fd\u8ba1\u7b97\u548c\u6e05\u6670\u7684\u8c03\u8bd5\u5de5\u5177\uff0c\u52a9\u529b\u5feb\u901f\u5f00\u53d1\u548c\u90e8\u7f72\u6df1\u5ea6\u5b66\u4e60\u6a21\u578b\u3002Keras \u80fd\u8f7b\u677e\u5e94\u5bf9\u56fe\u50cf\u5206\u7c7b\u3001\u81ea\u7136\u8bed\u8a00\u5904\u7406\u548c\u751f\u6210\u6a21\u578b\u7b49\u4efb\u52a1\uff0c\u662f\u6df1\u5ea6\u5b66\u4e60\u5f00\u53d1\u7684\u5f3a\u5927\u5de5\u5177\u3002<\/p>\n\n\n\n

\"Keras\"<\/a><\/figure>\n\n\n\n

Keras\u7684\u4e3b\u8981\u529f\u80fd<\/h2>\n\n\n\n
    \n
  • \u8de8\u6846\u67b6\u517c\u5bb9<\/strong>\uff1a\u652f\u6301 TensorFlow\u3001JAX \u548c PyTorch \u7b49\u591a\u79cd\u540e\u7aef\uff0c\u6a21\u578b\u80fd\u5728\u4e0d\u540c\u6846\u67b6\u4e4b\u95f4\u65e0\u7f1d\u5207\u6362\u3002<\/li>\n\n\n\n
  • \u7edf\u4e00 API<\/strong>\uff1aKeras \u7684 API \u59cb\u7ec8\u4fdd\u6301\u4e00\u81f4\uff0c\u964d\u4f4e\u5b66\u4e60\u6210\u672c\u3002<\/li>\n\n\n\n
  • \u6a21\u5757\u5316\u8bbe\u8ba1<\/strong>\uff1a\u6a21\u578b\u7ec4\u4ef6\u80fd\u50cf\u4e50\u9ad8\u79ef\u6728\u4e00\u6837\u81ea\u7531\u7ec4\u5408\uff0c\u65b9\u4fbf\u590d\u7528\u548c\u6269\u5c55\u3002<\/li>\n\n\n\n
  • \u9ad8\u6027\u80fd\u8ba1\u7b97<\/strong>\uff1a\u5229\u7528 JAX \u7684\u52a0\u901f\u80fd\u529b\uff0c\u63d0\u5347\u6a21\u578b\u8bad\u7ec3\u6548\u7387\u3002<\/li>\n\n\n\n
  • \u6613\u4e8e\u8c03\u8bd5<\/strong>\uff1a\u63d0\u4f9b\u6e05\u6670\u7684\u9519\u8bef\u4fe1\u606f\u548c\u8c03\u8bd5\u5de5\u5177\uff0c\u8ba9\u5f00\u53d1\u8fc7\u7a0b\u66f4\u987a\u7545\u3002<\/li>\n\n\n\n
  • \u5feb\u901f\u539f\u578b\u5f00\u53d1<\/strong>\uff1a\u4ece\u60f3\u6cd5\u5230\u5b9e\u9a8c\u9a8c\u8bc1\u53ea\u9700\u51e0\u5206\u949f\uff0c\u9002\u5408\u5feb\u901f\u8fed\u4ee3\u3002<\/li>\n\n\n\n
  • \u4e30\u5bcc\u7684\u9884\u8bad\u7ec3\u6a21\u578b<\/strong>\uff1a\u63d0\u4f9b VGG\u3001ResNet\u3001BERT \u7b49\u9884\u8bad\u7ec3\u6a21\u578b\uff0c\u65b9\u4fbf\u8fc1\u79fb\u5b66\u4e60\u3002<\/li>\n\n\n\n
  • \u7075\u6d3b\u7684\u6a21\u578b\u6784\u5efa\u65b9\u5f0f<\/strong>\uff1a\u652f\u6301 Sequential \u548c Functional API\uff0c\u9002\u5408\u4ece\u7b80\u5355\u5230\u590d\u6742\u7684\u6a21\u578b\u6784\u5efa\u3002<\/li>\n\n\n\n
  • \u4e00\u7ad9\u5f0f\u8bad\u7ec3\u4e0e\u8bc4\u4f30<\/strong>\uff1a\u63d0\u4f9b\u7f16\u8bd1\u3001\u8bad\u7ec3\u3001\u8bc4\u4f30\u548c\u9884\u6d4b\u7684\u5b8c\u6574\u6d41\u7a0b\u3002<\/li>\n\n\n\n
  • \u751f\u4ea7\u7ea7\u90e8\u7f72<\/strong>\uff1a\u6a21\u578b\u80fd\u5bfc\u51fa\u4e3a TensorFlow Lite\u3001ONNX \u7b49\u683c\u5f0f\uff0c\u652f\u6301\u591a\u5e73\u53f0\u90e8\u7f72\u3002<\/li>\n<\/ul>\n\n\n\n

    \u5982\u4f55\u4f7f\u7528Keras<\/h2>\n\n\n\n
      \n
    • \u5b89\u88c5 Keras<\/strong>\uff1aKeras \u662f\u72ec\u7acb\u7684\u6df1\u5ea6\u5b66\u4e60\u6846\u67b6\uff0c\u53ef\u901a\u8fc7 pip \u5b89\u88c5\u3002\u5982\u679c\u4f7f\u7528\u7684\u662f TensorFlow 2.x\uff0cKeras \u5df2\u7ecf\u96c6\u6210\u5728\u5176\u4e2d\uff0c\u80fd\u76f4\u63a5\u4f7f\u7528\u00a0tensorflow.keras<\/code>\u3002<\/li>\n\n\n\n
    • \u5bfc\u5165\u5fc5\u8981\u7684\u6a21\u5757<\/strong>\uff1a\u5728 Python \u811a\u672c\u6216 Jupyter Notebook \u4e2d\uff0c\u5bfc\u5165 Keras \u7684\u76f8\u5173\u6a21\u5757\uff0c\u5305\u62ec\u6a21\u578b\u6784\u5efa\u6a21\u5757\uff08\u5982\u00a0Sequential<\/code>\u00a0\u548c\u00a0Functional API<\/code>\uff09\u3001\u5c42\u6a21\u5757\uff08\u5982\u00a0Dense<\/code>\u3001Conv2D<\/code>\u00a0\u7b49\uff09\u53ca\u5176\u4ed6\u8f85\u52a9\u6a21\u5757\uff08\u5982\u00a0callbacks<\/code>\u00a0\u548c\u00a0preprocessing<\/code>\uff09\u3002<\/li>\n\n\n\n
    • \u6784\u5efa\u6a21\u578b<\/strong>\n
        \n
      • \u9009\u62e9\u5efa\u6a21\u65b9\u5f0f<\/strong>\uff1aKeras \u63d0\u4f9b\u4e24\u79cd\u4e3b\u8981\u7684\u5efa\u6a21\u65b9\u5f0f\uff1a\n
          \n
        • Sequential API<\/strong>\uff1a\u9002\u5408\u6784\u5efa\u7b80\u5355\u7684\u7ebf\u6027\u5806\u53e0\u6a21\u578b\u3002\u901a\u8fc7\u5c06\u5c42\u4f9d\u6b21\u6dfb\u52a0\u5230\u00a0Sequential<\/code>\u00a0\u6a21\u578b\u4e2d\u6784\u5efa\u7f51\u7edc\u3002<\/li>\n\n\n\n
        • Functional API<\/strong>\uff1a\u9002\u5408\u6784\u5efa\u590d\u6742\u7684\u6a21\u578b\uff0c\u5982\u591a\u8f93\u5165\u591a\u8f93\u51fa\u6a21\u578b\u3001\u6b8b\u5dee\u7f51\u7edc\u7b49\u3002\u901a\u8fc7\u5b9a\u4e49\u8f93\u5165\u548c\u8f93\u51fa\u6765\u6784\u5efa\u6a21\u578b\u3002<\/li>\n<\/ul>\n<\/li>\n\n\n\n
        • \u5b9a\u4e49\u6a21\u578b\u7ed3\u6784<\/strong>\uff1a\u6839\u636e\u4efb\u52a1\u9700\u6c42\uff0c\u9009\u62e9\u5408\u9002\u7684\u5c42\u7c7b\u578b\uff08\u5982\u5168\u8fde\u63a5\u5c42\u3001\u5377\u79ef\u5c42\u3001\u6c60\u5316\u5c42\u7b49\uff09\u6784\u5efa\u6a21\u578b\u3002<\/li>\n<\/ul>\n<\/li>\n\n\n\n
        • \u7f16\u8bd1\u6a21\u578b<\/strong>\uff1a\u5728\u8bad\u7ec3\u6a21\u578b\u4e4b\u524d\uff0c\u9700\u8981\u914d\u7f6e\u8bad\u7ec3\u8fc7\u7a0b\u3002\u5305\u62ec\u9009\u62e9\u4f18\u5316\u5668\uff08\u5982 Adam\u3001SGD \u7b49\uff09\u3001\u635f\u5931\u51fd\u6570\uff08\u5982\u4ea4\u53c9\u71b5\u635f\u5931\u3001\u5747\u65b9\u8bef\u5dee\u7b49\uff09\u548c\u8bc4\u4f30\u6307\u6807\uff08\u5982\u51c6\u786e\u7387\u3001\u53ec\u56de\u7387\u7b49\uff09\u3002\u7f16\u8bd1\u6a21\u578b\u662f\u5c06\u914d\u7f6e\u5e94\u7528\u5230\u6a21\u578b\u4e2d\u7684\u8fc7\u7a0b\u3002<\/li>\n\n\n\n
        • \u51c6\u5907\u6570\u636e<\/strong>\uff1a\u52a0\u8f7d\u548c\u9884\u5904\u7406\u6570\u636e\u662f\u8bad\u7ec3\u6a21\u578b\u7684\u91cd\u8981\u6b65\u9aa4\u3002\u9700\u8981\u5c06\u6570\u636e\u8f6c\u6362\u4e3a\u9002\u5408\u6a21\u578b\u8f93\u5165\u7684\u683c\u5f0f\uff0c\u8fdb\u884c\u5f52\u4e00\u5316\u3001\u6807\u51c6\u5316\u7b49\u9884\u5904\u7406\u64cd\u4f5c\u3002<\/li>\n\n\n\n
        • \u8bad\u7ec3\u6a21\u578b<\/strong>\uff1a\u7528\u00a0fit<\/code>\u00a0\u65b9\u6cd5\u8bad\u7ec3\u6a21\u578b\u3002\u9700\u8981\u6307\u5b9a\u8bad\u7ec3\u6570\u636e\u3001\u9a8c\u8bc1\u6570\u636e\u3001\u8bad\u7ec3\u8f6e\u6b21\uff08epochs\uff09\u548c\u6279\u6b21\u5927\u5c0f\uff08batch size\uff09\u7b49\u53c2\u6570\u3002\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\uff0c\u6a21\u578b\u6839\u636e\u635f\u5931\u51fd\u6570\u548c\u4f18\u5316\u5668\u9010\u6b65\u8c03\u6574\u6743\u91cd\uff0c\u4ee5\u6700\u5c0f\u5316\u635f\u5931\u3002<\/li>\n\n\n\n
        • \u8bc4\u4f30\u6a21\u578b<\/strong>\uff1a\u8bad\u7ec3\u5b8c\u6210\u540e\uff0c\u4f7f\u7528\u6d4b\u8bd5\u96c6\u8bc4\u4f30\u6a21\u578b\u7684\u6027\u80fd\u3002\u8bc4\u4f30\u6307\u6807\u901a\u5e38\u5305\u62ec\u51c6\u786e\u7387\u3001\u53ec\u56de\u7387\u3001F1 \u5206\u6570\u7b49\u3002\u901a\u8fc7\u8bc4\u4f30\u6a21\u578b\uff0c\u4e86\u89e3\u6a21\u578b\u5728\u672a\u89c1\u6570\u636e\u4e0a\u7684\u8868\u73b0\u3002<\/li>\n\n\n\n
        • \u4f7f\u7528\u6a21\u578b\u8fdb\u884c\u9884\u6d4b<\/strong>\uff1a\u8bad\u7ec3\u597d\u7684\u6a21\u578b\u7528\u5728\u5bf9\u65b0\u6570\u636e\u8fdb\u884c\u9884\u6d4b\u3002\u5c06\u65b0\u6570\u636e\u9884\u5904\u7406\u4e3a\u4e0e\u8bad\u7ec3\u6570\u636e\u76f8\u540c\u7684\u683c\u5f0f\uff0c\u7528\u6a21\u578b\u7684\u00a0predict<\/code>\u00a0\u65b9\u6cd5\u8fdb\u884c\u9884\u6d4b\u3002<\/li>\n\n\n\n
        • \u4fdd\u5b58\u548c\u52a0\u8f7d\u6a21\u578b<\/strong>\uff1a\u4e3a\u65b9\u4fbf\u540e\u7eed\u4f7f\u7528\uff0c\u5c06\u8bad\u7ec3\u597d\u7684\u6a21\u578b\u4fdd\u5b58\u5230\u78c1\u76d8\u3002Keras \u63d0\u4f9b\u4fdd\u5b58\u548c\u52a0\u8f7d\u6a21\u578b\u7684\u529f\u80fd\uff0c\u652f\u6301\u4fdd\u5b58\u6a21\u578b\u7684\u7ed3\u6784\u3001\u6743\u91cd\u548c\u8bad\u7ec3\u914d\u7f6e\u7b49\u4fe1\u606f\u3002<\/li>\n<\/ul>\n\n\n\n

          Keras\u7684\u5b98\u7f51\u5730\u5740<\/h2>\n\n\n\n
            \n
          • \u5b98\u7f51\u5730\u5740<\/strong>\uff1ahttps:\/\/keras.io\/<\/li>\n\n\n\n
          • GitHub\u4ed3\u5e93<\/strong>\uff1ahttps:\/\/github.com\/keras-team<\/li>\n<\/ul>\n\n\n\n

            Keras\u7684\u5e94\u7528\u573a\u666f<\/h2>\n\n\n\n
              \n
            • \u56fe\u50cf\u5206\u7c7b<\/strong>\uff1a\u7528\u5728\u8bc6\u522b\u56fe\u50cf\u4e2d\u7684\u7269\u4f53\u6216\u573a\u666f\uff0c\u5982 CIFAR-10 \u6570\u636e\u96c6\u7684\u5206\u7c7b\u4efb\u52a1\u3002<\/li>\n\n\n\n
            • \u81ea\u7136\u8bed\u8a00\u5904\u7406<\/strong>\uff1a\u652f\u6301\u5904\u7406\u6587\u672c\u6570\u636e\uff0c\u5305\u62ec\u6587\u672c\u5206\u7c7b\u3001\u60c5\u611f\u5206\u6790\u3001\u673a\u5668\u7ffb\u8bd1\u7b49\u3002<\/li>\n\n\n\n
            • \u63a8\u8350\u7cfb\u7edf<\/strong>\uff1a\u6784\u5efa\u7528\u6237\u548c\u7269\u54c1\u4e4b\u95f4\u7684\u5173\u7cfb\u6a21\u578b\uff0c\u9884\u6d4b\u7528\u6237\u5bf9\u7269\u54c1\u7684\u8bc4\u5206\u6216\u504f\u597d\u3002<\/li>\n\n\n\n
            • \u751f\u6210\u5bf9\u6297\u7f51\u7edc\uff08GAN\uff09<\/strong>\uff1a\u7528\u5728\u751f\u6210\u56fe\u50cf\u3001\u6587\u672c\u7b49\u6570\u636e\uff0c\u5982\u751f\u6210\u903c\u771f\u7684\u56fe\u50cf\u6216\u521b\u610f\u6587\u672c\u3002<\/li>\n\n\n\n
            • \u8fc1\u79fb\u5b66\u4e60<\/strong>\uff1a\u7528\u9884\u8bad\u7ec3\u6a21\u578b\u89e3\u51b3\u7279\u5b9a\u4efb\u52a1\uff0c\u5982\u4f7f\u7528 ResNet \u8fdb\u884c\u56fe\u50cf\u8bc6\u522b\u3002<\/li>\n<\/ul>\n","protected":false},"excerpt":{"rendered":"

              Keras Python\u7248\u672c\u7684TensorFlow\u6df1\u5ea6\u5b66\u4e60API Keras\u662f\u4ec0\u4e48 Keras \u662f\u5f00\u6e90\u7684\u6df1\u5ea6\u5b66 […]<\/p>\n","protected":false},"author":1,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"footnotes":""},"class_list":["post-2567","page","type-page","status-publish","hentry"],"_links":{"self":[{"href":"https:\/\/ai.ziyuanzz.online\/index.php\/wp-json\/wp\/v2\/pages\/2567","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/ai.ziyuanzz.online\/index.php\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/ai.ziyuanzz.online\/index.php\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/ai.ziyuanzz.online\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/ai.ziyuanzz.online\/index.php\/wp-json\/wp\/v2\/comments?post=2567"}],"version-history":[{"count":0,"href":"https:\/\/ai.ziyuanzz.online\/index.php\/wp-json\/wp\/v2\/pages\/2567\/revisions"}],"wp:attachment":[{"href":"https:\/\/ai.ziyuanzz.online\/index.php\/wp-json\/wp\/v2\/media?parent=2567"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}