{"id":28019,"date":"2025-03-10T00:40:25","date_gmt":"2025-03-09T16:40:25","guid":{"rendered":"https:\/\/www.aisharenet.com\/?p=28019"},"modified":"2025-03-10T00:40:25","modified_gmt":"2025-03-09T16:40:25","slug":"litserve","status":"publish","type":"post","link":"https:\/\/www.kdjingpai.com\/de\/litserve\/","title":{"rendered":"LitServe\uff1a\u5feb\u901f\u90e8\u7f72\u4f01\u4e1a\u7ea7\u901a\u7528AI\u6a21\u578b\u63a8\u7406\u670d\u52a1"},"content":{"rendered":"<p>LitServe \u662f <a href=\"https:\/\/www.kdjingpai.com\/de\/lightning\/\">Lightning<\/a> AI \u63a8\u51fa\u7684\u4e00\u6b3e\u5f00\u6e90 AI \u6a21\u578b\u670d\u52a1\u5f15\u64ce\uff0c\u57fa\u4e8e FastAPI \u6784\u5efa\uff0c\u4e13\u6ce8\u4e8e\u5feb\u901f\u90e8\u7f72\u901a\u7528 AI \u6a21\u578b\u7684\u63a8\u7406\u670d\u52a1\u3002\u5b83\u652f\u6301\u4ece\u5927\u578b\u8bed\u8a00\u6a21\u578b\uff08LLM\uff09\u3001\u89c6\u89c9\u6a21\u578b\u3001\u97f3\u9891\u6a21\u578b\u5230\u7ecf\u5178\u673a\u5668\u5b66\u4e60\u6a21\u578b\u7684\u5e7f\u6cdb\u573a\u666f\uff0c\u63d0\u4f9b\u6279\u91cf\u5904\u7406\u3001\u6d41\u5f0f\u4f20\u8f93\u548c GPU \u81ea\u52a8\u6269\u5c55\u7b49\u529f\u80fd\uff0c\u6027\u80fd\u6bd4\u666e\u901a FastAPI \u63d0\u5347\u81f3\u5c11 2 \u500d\u3002LitServe \u6613\u4e8e\u4f7f\u7528\u4e14\u9ad8\u5ea6\u7075\u6d3b\uff0c\u7528\u6237\u53ef\u81ea\u6258\u7ba1\u6216\u901a\u8fc7 Lightning Studios \u5b9e\u73b0\u5b8c\u5168\u6258\u7ba1\uff0c\u9002\u7528\u4e8e\u7814\u7a76\u4eba\u5458\u3001\u5f00\u53d1\u8005\u548c\u4f01\u4e1a\u5feb\u901f\u6784\u5efa\u9ad8\u6548\u7684\u6a21\u578b\u63a8\u7406 API\u3002\u5b98\u65b9\u5f3a\u8c03\u5176\u4f01\u4e1a\u7ea7\u7279\u6027\uff0c\u5982\u5b89\u5168\u6027\u3001\u53ef\u6269\u5c55\u6027\u548c\u9ad8\u53ef\u7528\u6027\uff0c\u786e\u4fdd\u751f\u4ea7\u73af\u5883\u5f00\u7bb1\u5373\u7528\u3002<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter  wp-image-28020\" title=\"LitServe\uff1a\u5feb\u901f\u90e8\u7f72\u4f01\u4e1a\u7ea7\u901a\u7528AI\u6a21\u578b\u63a8\u7406\u670d\u52a1-1\" src=\"https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2025\/03\/4809e3aa10cbf68.png\" alt=\"LitServe\uff1a\u5feb\u901f\u90e8\u7f72\u4f01\u4e1a\u7ea7\u901a\u7528AI\u6a21\u578b\u63a8\u7406\u670d\u52a1-1\" width=\"668\" height=\"295\" srcset=\"https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2025\/03\/4809e3aa10cbf68.png 835w, https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2025\/03\/4809e3aa10cbf68-768x339.png 768w, https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2025\/03\/4809e3aa10cbf68-18x8.png 18w\" sizes=\"auto, (max-width: 668px) 100vw, 668px\" \/><\/p>\n<p>&nbsp;<\/p>\n<h2>\u529f\u80fd\u5217\u8868<\/h2>\n<ul>\n<li><strong>\u5feb\u901f\u90e8\u7f72\u63a8\u7406\u670d\u52a1<\/strong>\uff1a\u652f\u6301 PyTorch\u3001JAX\u3001TensorFlow \u7b49\u6846\u67b6\u7684\u6a21\u578b\u5feb\u901f\u8f6c\u4e3a API\u3002<\/li>\n<li><strong>\u6279\u91cf\u5904\u7406<\/strong>\uff1a\u5408\u5e76\u591a\u4e2a\u63a8\u7406\u8bf7\u6c42\u6210\u6279\u6b21\u5904\u7406\uff0c\u63d0\u5347\u541e\u5410\u91cf\u3002<\/li>\n<li><strong>\u6d41\u5f0f\u4f20\u8f93<\/strong>\uff1a\u652f\u6301\u5b9e\u65f6\u63a8\u7406\u7ed3\u679c\u6d41\u8f93\u51fa\uff0c\u9002\u5408\u8fde\u7eed\u54cd\u5e94\u573a\u666f\u3002<\/li>\n<li><strong>GPU \u81ea\u52a8\u6269\u5c55<\/strong>\uff1a\u6839\u636e\u63a8\u7406\u8d1f\u8f7d\u52a8\u6001\u8c03\u6574 GPU \u8d44\u6e90\uff0c\u4f18\u5316\u6027\u80fd\u3002<\/li>\n<li><strong>\u590d\u5408 AI \u7cfb\u7edf<\/strong>\uff1a\u5141\u8bb8\u591a\u4e2a\u6a21\u578b\u534f\u540c\u63a8\u7406\uff0c\u6784\u5efa\u590d\u6742\u670d\u52a1\u3002<\/li>\n<li><strong>\u81ea\u6258\u7ba1\u4e0e\u4e91\u6258\u7ba1<\/strong>\uff1a\u652f\u6301\u672c\u5730\u90e8\u7f72\u6216\u901a\u8fc7 Lightning Studios \u4e91\u7aef\u7ba1\u7406\u3002<\/li>\n<li><strong>\u4e0e vLLM \u96c6\u6210<\/strong>\uff1a\u4f18\u5316\u5927\u578b\u8bed\u8a00\u6a21\u578b\u7684\u63a8\u7406\u6027\u80fd\u3002<\/li>\n<li><strong>OpenAPI \u517c\u5bb9<\/strong>\uff1a\u81ea\u52a8\u751f\u6210\u6807\u51c6 API \u6587\u6863\uff0c\u4fbf\u4e8e\u6d4b\u8bd5\u548c\u96c6\u6210\u3002<\/li>\n<li><strong>\u5168\u6a21\u578b\u652f\u6301<\/strong>\uff1a\u8986\u76d6 LLM\u3001\u89c6\u89c9\u3001\u97f3\u9891\u3001\u5d4c\u5165\u7b49\u5404\u79cd\u6a21\u578b\u7684\u63a8\u7406\u9700\u6c42\u3002<\/li>\n<li><strong>\u670d\u52a1\u5668\u4f18\u5316<\/strong>\uff1a\u63d0\u4f9b\u591a\u5de5\u4f5c\u8fdb\u7a0b\u5904\u7406\uff0c\u63a8\u7406\u901f\u5ea6\u6bd4 FastAPI \u5feb 2 \u500d\u4ee5\u4e0a\u3002<\/li>\n<\/ul>\n<p>&nbsp;<\/p>\n<h2>\u4f7f\u7528\u5e2e\u52a9<\/h2>\n<h3>\u5b89\u88c5\u6d41\u7a0b<\/h3>\n<p>LitServe \u7684\u5b89\u88c5\u7b80\u5355\uff0c\u901a\u8fc7 Python \u7684\u00a0<code>pip<\/code>\u00a0\u5de5\u5177\u5373\u53ef\u5b8c\u6210\u3002\u4ee5\u4e0b\u662f\u8be6\u7ec6\u6b65\u9aa4\uff1a<\/p>\n<h4>1. \u51c6\u5907\u73af\u5883<\/h4>\n<p>\u786e\u4fdd\u7cfb\u7edf\u5b89\u88c5\u4e86 Python 3.8 \u6216\u4ee5\u4e0a\u7248\u672c\uff0c\u5efa\u8bae\u4f7f\u7528\u865a\u62df\u73af\u5883\uff1a<\/p>\n<pre><code>python -m venv venv\r\nsource venv\/bin\/activate  # Linux\/Mac\r\nvenv\\Scripts\\activate     # Windows\r\n<\/code><\/pre>\n<h4>2. \u5b89\u88c5 LitServe<\/h4>\n<p>\u8fd0\u884c\u4ee5\u4e0b\u547d\u4ee4\u5b89\u88c5\u7a33\u5b9a\u7248\uff1a<\/p>\n<pre><code>pip install litserve\r\n<\/code><\/pre>\n<p>\u82e5\u9700\u6700\u65b0\u529f\u80fd\uff0c\u53ef\u5b89\u88c5\u5f00\u53d1\u7248\uff1a<\/p>\n<pre><code>pip install git+https:\/\/github.com\/Lightning-AI\/litserve.git@main\r\n<\/code><\/pre>\n<h4>3. \u68c0\u67e5\u5b89\u88c5<\/h4>\n<p>\u9a8c\u8bc1\u662f\u5426\u6210\u529f\uff1a<\/p>\n<pre><code>python -c \"import litserve; print(litserve.__version__)\"\r\n<\/code><\/pre>\n<p>\u6210\u529f\u8f93\u51fa\u7248\u672c\u53f7\u5373\u5b8c\u6210\u5b89\u88c5\u3002<\/p>\n<h4>4. \u53ef\u9009\u4f9d\u8d56<\/h4>\n<p>\u82e5\u9700 GPU \u652f\u6301\uff0c\u5b89\u88c5\u5bf9\u5e94\u6846\u67b6\u7684 GPU \u7248\u672c\uff0c\u4f8b\u5982\uff1a<\/p>\n<pre><code>pip install torch torchvision --index-url https:\/\/download.pytorch.org\/whl\/cu121\r\n<\/code><\/pre>\n<h3>\u5982\u4f55\u4f7f\u7528 LitServe<\/h3>\n<p>LitServe \u901a\u8fc7\u7b80\u6d01\u4ee3\u7801\u5c06 AI \u6a21\u578b\u8f6c\u4e3a\u63a8\u7406\u670d\u52a1\u3002\u4ee5\u4e0b\u662f\u8be6\u7ec6\u64cd\u4f5c\u6d41\u7a0b\uff1a<\/p>\n<h4>1. \u521b\u5efa\u7b80\u5355\u63a8\u7406\u670d\u52a1<\/h4>\n<p>\u4ee5\u4e0b\u662f\u4e00\u4e2a\u590d\u5408\u63a8\u7406\u670d\u52a1\u793a\u4f8b\uff0c\u5305\u542b\u4e24\u4e2a\u6a21\u578b\uff1a<\/p>\n<pre><code>import litserve as ls\r\nclass SimpleLitAPI(ls.LitAPI):\r\ndef setup(self, device):\r\n# \u521d\u59cb\u5316\uff0c\u52a0\u8f7d\u6a21\u578b\u6216\u6570\u636e\r\nself.model1 = lambda x: x ** 2  # \u5e73\u65b9\u6a21\u578b\r\nself.model2 = lambda x: x ** 3  # \u7acb\u65b9\u6a21\u578b\r\ndef decode_request(self, request):\r\n# \u89e3\u6790\u8bf7\u6c42\u6570\u636e\r\nreturn request[\"input\"]\r\ndef predict(self, x):\r\n# \u590d\u5408\u63a8\u7406\r\nsquared = self.model1(x)\r\ncubed = self.model2(x)\r\nreturn squared + cubed\r\ndef encode_response(self, output):\r\n# \u683c\u5f0f\u5316\u63a8\u7406\u7ed3\u679c\r\nreturn {\"output\": output}\r\nif __name__ == \"__main__\":\r\nserver = ls.LitServer(SimpleLitAPI(), accelerator=\"auto\")\r\nserver.run(port=8000)\r\n<\/code><\/pre>\n<ul>\n<li><strong>\u8fd0\u884c<\/strong>\uff1a\u4fdd\u5b58\u4e3a\u00a0<code>server.py<\/code>\uff0c\u6267\u884c\u00a0<code>python server.py<\/code>\u3002<\/li>\n<li><strong>\u6d4b\u8bd5<\/strong>\uff1a\u7528\u00a0<code>curl<\/code>\u00a0\u53d1\u9001\u63a8\u7406\u8bf7\u6c42\uff1a\n<pre><code>curl -X POST \"http:\/\/127.0.0.1:8000\/predict\" -H \"Content-Type: application\/json\" -d '{\"input\": 4.0}'\r\n<\/code><\/pre>\n<p>\u8f93\u51fa\uff1a<code>{\"output\": 80.0}<\/code>\uff0816 + 64\uff09\u3002<\/li>\n<\/ul>\n<h4>2. \u542f\u7528\u6279\u91cf\u63a8\u7406<\/h4>\n<p>\u4fee\u6539\u4ee3\u7801\u652f\u6301\u6279\u91cf\u5904\u7406\uff1a<\/p>\n<pre><code>server = ls.LitServer(SimpleLitAPI(), max_batch_size=4, accelerator=\"auto\")\r\n<\/code><\/pre>\n<ul>\n<li><strong>\u64cd\u4f5c\u8bf4\u660e<\/strong>\uff1a<code>max_batch_size=4<\/code>\u00a0\u8868\u793a\u6700\u591a\u540c\u65f6\u5904\u7406 4 \u4e2a\u63a8\u7406\u8bf7\u6c42\uff0c\u81ea\u52a8\u5408\u5e76\u63d0\u5347\u6548\u7387\u3002<\/li>\n<li><strong>\u6d4b\u8bd5\u65b9\u6cd5<\/strong>\uff1a\u591a\u6b21\u53d1\u9001\u8bf7\u6c42\uff0c\u89c2\u5bdf\u541e\u5410\u91cf\u63d0\u5347\uff1a\n<pre><code>curl -X POST \"http:\/\/127.0.0.1:8000\/predict\" -H \"Content-Type: application\/json\" -d '{\"input\": 5.0}'\r\n<\/code><\/pre>\n<\/li>\n<\/ul>\n<h4>3. \u914d\u7f6e\u6d41\u5f0f\u63a8\u7406<\/h4>\n<p>\u9002\u7528\u4e8e\u5b9e\u65f6\u63a8\u7406\u573a\u666f\uff1a<\/p>\n<pre><code>class StreamLitAPI(ls.LitAPI):\r\ndef setup(self, device):\r\nself.model = lambda x: [x * i for i in range(5)]\r\ndef decode_request(self, request):\r\nreturn request[\"input\"]\r\ndef predict(self, x):\r\nfor result in self.model(x):\r\nyield result\r\ndef encode_response(self, output):\r\nreturn {\"output\": output}\r\nserver = ls.LitServer(StreamLitAPI(), stream=True, accelerator=\"auto\")\r\nserver.run(port=8000)\r\n<\/code><\/pre>\n<ul>\n<li><strong>\u64cd\u4f5c\u8bf4\u660e<\/strong>\uff1a<code>stream=True<\/code>\u00a0\u542f\u7528\u6d41\u5f0f\u63a8\u7406\uff0c<code>predict<\/code>\u00a0\u4f7f\u7528\u00a0<code>yield<\/code>\u00a0\u9010\u4e2a\u8fd4\u56de\u7ed3\u679c\u3002<\/li>\n<li><strong>\u6d4b\u8bd5\u65b9\u6cd5<\/strong>\uff1a\u4f7f\u7528\u652f\u6301\u6d41\u5f0f\u54cd\u5e94\u7684\u5ba2\u6237\u7aef\uff1a\n<pre><code>curl --no-buffer -X POST \"http:\/\/127.0.0.1:8000\/predict\" -H \"Content-Type: application\/json\" -d '{\"input\": 2}'\r\n<\/code><\/pre>\n<\/li>\n<\/ul>\n<h4>4. GPU \u81ea\u52a8\u6269\u5c55<\/h4>\n<p>\u82e5\u6709 GPU\uff0cLitServe \u81ea\u52a8\u4f18\u5316\u63a8\u7406\uff1a<\/p>\n<ul>\n<li><strong>\u64cd\u4f5c\u8bf4\u660e<\/strong>\uff1a<code>accelerator=\"auto\"<\/code>\u00a0\u68c0\u6d4b\u5e76\u4f18\u5148\u4f7f\u7528 GPU\u3002<\/li>\n<li><strong>\u9a8c\u8bc1<\/strong>\uff1a\u8fd0\u884c\u540e\u67e5\u770b\u65e5\u5fd7\uff0c\u786e\u8ba4 GPU \u4f7f\u7528\u3002<\/li>\n<li><strong>\u73af\u5883\u8981\u6c42<\/strong>\uff1a\u786e\u4fdd\u5b89\u88c5 GPU \u7248\u672c\u6846\u67b6\uff08\u5982 PyTorch\uff09\u3002<\/li>\n<\/ul>\n<h4>5. \u90e8\u7f72\u590d\u6742\u6a21\u578b\u63a8\u7406\uff08\u4ee5 BERT \u4e3a\u4f8b\uff09<\/h4>\n<p>\u90e8\u7f72 Hugging Face \u7684 BERT \u6a21\u578b\u63a8\u7406\u670d\u52a1\uff1a<\/p>\n<pre><code>from transformers import BertTokenizer, BertModel\r\nimport litserve as ls\r\nclass BertLitAPI(ls.LitAPI):\r\ndef setup(self, device):\r\nself.tokenizer = BertTokenizer.from_pretrained(\"bert-base-uncased\")\r\nself.model = BertModel.from_pretrained(\"bert-base-uncased\").to(device)\r\ndef decode_request(self, request):\r\nreturn request[\"text\"]\r\ndef predict(self, text):\r\ninputs = self.tokenizer(text, return_tensors=\"pt\").to(self.model.device)\r\noutputs = self.model(**inputs)\r\nreturn outputs.last_hidden_state.mean(dim=1).tolist()\r\ndef encode_response(self, output):\r\nreturn {\"embedding\": output}\r\nserver = ls.LitServer(BertLitAPI(), accelerator=\"auto\")\r\nserver.run(port=8000)\r\n<\/code><\/pre>\n<ul>\n<li><strong>\u8fd0\u884c<\/strong>\uff1a\u6267\u884c\u811a\u672c\u540e\uff0c\u8bbf\u95ee\u00a0<code>http:\/\/127.0.0.1:8000\/predict<\/code>\u3002<\/li>\n<li><strong>\u6d4b\u8bd5<\/strong>\uff1a\n<pre><code>curl -X POST \"http:\/\/127.0.0.1:8000\/predict\" -H \"Content-Type: application\/json\" -d '{\"text\": \"Hello, world!\"}'\r\n<\/code><\/pre>\n<\/li>\n<\/ul>\n<h4>6. \u96c6\u6210 vLLM \u90e8\u7f72 LLM \u63a8\u7406<\/h4>\n<p>\u4e3a\u5927\u578b\u8bed\u8a00\u6a21\u578b\u63d0\u4f9b\u9ad8\u6548\u63a8\u7406\uff1a<\/p>\n<pre><code>import litserve as ls\r\nfrom <a href=\"https:\/\/www.kdjingpai.com\/de\/vllm\/\">vllm<\/a> import LLM\r\nclass LLMLitAPI(ls.LitAPI):\r\ndef setup(self, device):\r\nself.model = LLM(model=\"meta-llama\/Llama-3.2-1B\", dtype=\"float16\")\r\ndef decode_request(self, request):\r\nreturn request[\"prompt\"]\r\ndef predict(self, prompt):\r\noutputs = self.model.generate(prompt, max_tokens=50)\r\nreturn outputs[0].outputs[0].text\r\ndef encode_response(self, output):\r\nreturn {\"response\": output}\r\nserver = ls.LitServer(LLMLitAPI(), accelerator=\"auto\")\r\nserver.run(port=8000)\r\n<\/code><\/pre>\n<ul>\n<li><strong>\u5b89\u88c5 vLLM<\/strong>\uff1a<code>pip install vllm<\/code>\u3002<\/li>\n<li><strong>\u6d4b\u8bd5<\/strong>\uff1a\n<pre><code>curl -X POST \"http:\/\/127.0.0.1:8000\/predict\" -H \"Content-Type: application\/json\" -d '{\"prompt\": \"What is AI?\"}'\r\n<\/code><\/pre>\n<\/li>\n<\/ul>\n<h4>7. \u67e5\u770b API \u6587\u6863<\/h4>\n<ul>\n<li><strong>\u64cd\u4f5c\u8bf4\u660e<\/strong>\uff1a\u8bbf\u95ee\u00a0<code>http:\/\/127.0.0.1:8000\/docs<\/code>\uff0c\u4ea4\u4e92\u5f0f\u6d4b\u8bd5\u63a8\u7406\u670d\u52a1\u3002<\/li>\n<li><strong>\u529f\u80fd\u63d0\u793a<\/strong>\uff1a\u57fa\u4e8e OpenAPI \u6807\u51c6\uff0c\u5305\u542b\u6240\u6709\u7aef\u70b9\u8be6\u60c5\u3002<\/li>\n<\/ul>\n<h4>8. \u6258\u7ba1\u9009\u9879<\/h4>\n<ul>\n<li><strong>\u81ea\u6258\u7ba1<\/strong>\uff1a\u672c\u5730\u6216\u670d\u52a1\u5668\u8fd0\u884c\u4ee3\u7801\u3002<\/li>\n<li><strong>\u4e91\u6258\u7ba1<\/strong>\uff1a\u901a\u8fc7 Lightning Studios \u90e8\u7f72\uff0c\u9700\u6ce8\u518c\u8d26\u6237\uff0c\u63d0\u4f9b\u8d1f\u8f7d\u5747\u8861\u3001\u81ea\u52a8\u6269\u5c55\u7b49\u529f\u80fd\u3002<\/li>\n<\/ul>\n<h3>\u64cd\u4f5c\u5c0f\u8d34\u58eb<\/h3>\n<ul>\n<li><strong>\u8c03\u8bd5<\/strong>\uff1a\u8bbe\u7f6e\u00a0<code>timeout=60<\/code>\u00a0\u907f\u514d\u63a8\u7406\u8d85\u65f6\u3002<\/li>\n<li><strong>\u65e5\u5fd7<\/strong>\uff1a\u542f\u52a8\u65f6\u67e5\u770b\u7ec8\u7aef\u65e5\u5fd7\uff0c\u6392\u67e5\u95ee\u9898\u3002<\/li>\n<li><strong>\u4f18\u5316<\/strong>\uff1a\u53c2\u8003\u5b98\u65b9\u6587\u6863\u542f\u7528\u8ba4\u8bc1\u3001Docker \u90e8\u7f72\u7b49\u9ad8\u7ea7\u529f\u80fd\u3002<\/li>\n<\/ul>\n<p>LitServe \u901a\u8fc7\u5feb\u901f\u90e8\u7f72\u548c\u4f18\u5316\u63a8\u7406\u670d\u52a1\uff0c\u652f\u6301\u4ece\u539f\u578b\u9a8c\u8bc1\u5230\u4f01\u4e1a\u7ea7\u5e94\u7528\u7684\u5168\u6d41\u7a0b\u9700\u6c42\u3002<\/p>\n","protected":false},"excerpt":{"rendered":"<p>LitServe \u662f Lightning AI \u63a8\u51fa\u7684\u4e00\u6b3e\u5f00\u6e90 AI \u6a21\u578b\u670d\u52a1\u5f15\u64ce\uff0c\u57fa\u4e8e FastAPI \u6784\u5efa\uff0c\u4e13\u6ce8\u4e8e\u5feb\u901f\u90e8\u7f72\u901a\u7528 AI \u6a21\u578b\u7684\u63a8\u7406\u670d\u52a1\u3002\u5b83\u652f\u6301\u4ece\u5927\u578b\u8bed\u8a00\u6a21\u578b\uff08LLM\uff09\u3001\u89c6\u89c9\u6a21\u578b\u3001\u97f3\u9891\u6a21\u578b\u5230\u7ecf\u5178\u673a\u5668\u5b66\u4e60\u6a21\u578b\u7684\u5e7f\u6cdb\u573a\u666f\uff0c\u63d0\u4f9b\u6279&#8230;<\/p>\n","protected":false},"author":1,"featured_media":61999,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[20],"tags":[230,232],"class_list":["post-28019","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-tool","tag-aikaiyuanxiangmu","tag-bendebushukaiyuanba"],"_links":{"self":[{"href":"https:\/\/www.kdjingpai.com\/de\/wp-json\/wp\/v2\/posts\/28019","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.kdjingpai.com\/de\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.kdjingpai.com\/de\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.kdjingpai.com\/de\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.kdjingpai.com\/de\/wp-json\/wp\/v2\/comments?post=28019"}],"version-history":[{"count":0,"href":"https:\/\/www.kdjingpai.com\/de\/wp-json\/wp\/v2\/posts\/28019\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.kdjingpai.com\/de\/wp-json\/wp\/v2\/media\/61999"}],"wp:attachment":[{"href":"https:\/\/www.kdjingpai.com\/de\/wp-json\/wp\/v2\/media?parent=28019"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.kdjingpai.com\/de\/wp-json\/wp\/v2\/categories?post=28019"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.kdjingpai.com\/de\/wp-json\/wp\/v2\/tags?post=28019"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}