{"id":43497,"date":"2023-09-12T19:02:00","date_gmt":"2023-09-12T11:02:00","guid":{"rendered":"https:\/\/www.ioiotimes.com\/?p=43497"},"modified":"2023-09-13T00:10:23","modified_gmt":"2023-09-12T16:10:23","slug":"nvidia-grace-hopper%e8%b6%85%e7%b4%9a%e6%99%b6%e7%89%87%e5%9c%a8mlperf%e6%8e%a8%e8%ab%96%e5%9f%ba%e6%ba%96%e6%b8%ac%e8%a9%a6%e4%b8%ad%e5%8f%96%e5%be%97%e5%8d%93%e8%b6%8a%e6%88%90%e6%9e%9c","status":"publish","type":"post","link":"https:\/\/www.ioiotimes.com\/?p=43497","title":{"rendered":"NVIDIA Grace Hopper\u8d85\u7d1a\u6676\u7247\u5728MLPerf\u63a8\u8ad6\u57fa\u6e96\u6e2c\u8a66\u4e2d\u53d6\u5f97\u5353\u8d8a\u6210\u679c"},"content":{"rendered":"\n

NVIDIA GH200\u3001H100\u548cL4 GPU\u4ee5\u53caJetson Orin\u7cfb\u7d71\u6a21\u7d44\u5728\u5f9e\u96f2\u7aef\u5230\u7db2\u8def\u908a\u7de3\u7684\u751f\u7522\u74b0\u5883\u4e2d\u904b\u884c\u4eba\u5de5\u667a\u6167\u65b9\u9762\u8868\u73fe\u51fa\u9818\u5148\u7684\u6548\u80fd<\/h3>\n\n\n\n

NVIDIA GH200 Grace Hopper\u8d85\u7d1a\u6676\u7247<\/a> \u9996\u6b21\u4eae\u76f8\u65bc MLPerf \u7522\u696d\u57fa\u6e96\u6e2c\u8a66\u4e2d\uff0c\u5728\u6240\u6709\u4eba\u5de5\u667a\u6167\u63a8\u8ad6\u52a0\u901f\u5668\u6e2c\u8a66\u4e2d\u5747\u8868\u73fe\u512a\u7570\uff0c\u9032\u4e00\u6b65\u64f4\u5c55\u4e86 NVIDIA H100 Tensor Core GPU<\/a> \u7684\u9818\u5148\u6548\u80fd\u3002<\/p>\n\n\n\n

\u9019\u4e9b\u7d50\u679c\u9084\u5c55\u793a\u4e86 NVIDIA \u4eba\u5de5\u667a\u6167\u5e73\u53f0\u5728\u5f9e\u96f2\u7aef\u5230\u7db2\u8def\u908a\u7de3\u7684\u5353\u8d8a\u6027\u80fd\u548c\u591a\u529f\u80fd\u6027\u3002<\/p>\n\n\n\n

NVIDIA \u53e6\u5916\u5ba3\u5e03\u63a8\u51fa\u63a8\u8ad6\u8edf\u9ad4\uff0c\u80fd\u8b93\u4f7f\u7528\u8005\u5728\u6548\u80fd\u3001\u80fd\u6e90\u6548\u7387\u548c\u7e3d\u6301\u6709\u6210\u672c\u65b9\u9762\u4e0a\u5f97\u5230\u986f\u8457\u7684\u63d0\u5347\u3002<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

<\/p>\n\n\n\n

GH200<\/strong>\u8d85\u7d1a\u6676\u7247\u5728<\/strong>MLPerf<\/strong>\u57fa\u6e96\u6e2c\u8a66\u4e2d\u8868\u73fe\u512a\u7570<\/strong><\/strong><\/h3>\n\n\n\n

GH200 \u9023\u7d50 Hopper GPU \u548c Grace CPU \u6210\u70ba\u4e00\u500b\u8d85\u7d1a\u6676\u7247\u3002\u9019\u500b\u7d44\u5408\u80fd\u63d0\u4f9b\u66f4\u591a\u8a18\u61b6\u9ad4\u3001\u983b\u5bec\uff0c\u4ee5\u53ca\u80fd\u5728 CPU \u548c GPU \u4e4b\u9593\u81ea\u52d5\u8abf\u7bc0\u96fb\u529b\uff0c\u4ee5\u6700\u4f73\u5316\u8868\u73fe\u3002<\/p>\n\n\n\n

\u6b64\u5916\uff0c\u914d\u5099 8 \u500b H100 GPU \u7684 HGX H100 \u7cfb\u7d71<\/a>\u5728\u672c\u8f2a\u6bcf\u500b MLPerf \u63a8\u8ad6\u6e2c\u8a66\u4e2d\u63d0\u4f9b\u4e86\u6700\u9ad8\u7684\u541e\u5410\u91cf\u3002<\/p>\n\n\n\n

Grace Hopper \u8d85\u7d1a\u6676\u7247\u548c H100 GPU \u5728\u6240\u6709 MLPerf \u7684\u8cc7\u6599\u4e2d\u5fc3\u6e2c\u8a66\u4e2d\u8655\u65bc\u9818\u5148\u5730\u4f4d\uff0c\u5305\u62ec\u96fb\u8166\u8996\u89ba\u63a8\u8ad6\u3001\u8a9e\u97f3\u8b58\u5225\u548c\u91ab\u5b78\u6210\u50cf\uff0c\u4ee5\u53ca\u8981\u6c42\u66f4\u9ad8\u7684\u63a8\u85a6\u7cfb\u7d71\u61c9\u7528\u6848\u4f8b\u548c\u751f\u6210\u5f0f\u4eba\u5de5\u667a\u6167<\/a>\u4e2d\u4f7f\u7528\u7684\u5927\u578b\u8a9e\u8a00\u6a21\u578b (LLMs<\/a>)\u3002<\/p>\n\n\n\n

\u7e3d\u9ad4\u4f86\u8aaa\uff0c\u9019\u6b21\u6e2c\u8a66\u7d50\u679c\u5ef6\u7e8c\u4e86 NVIDIA \u81ea 2018 \u5e74 MLPerf \u57fa\u6e96\u63a8\u51fa\u4ee5\u4f86\uff0c\u5728\u6bcf\u8f2a\u4eba\u5de5\u667a\u6167\u8a13\u7df4\u548c\u63a8\u8ad6\u65b9\u9762\u6548\u80fd\u9818\u5148\u7684\u8a18\u9304<\/a>\u3002<\/p>\n\n\n\n

\u6700\u65b0\u7684 MLPerf \u6e2c\u8a66\u4e2d\u5305\u62ec\u5c0d\u63a8\u85a6\u7cfb\u7d71\u7684\u66f4\u65b0\u6e2c\u8a66\uff0c\u4ee5\u53ca\u9996\u6b21\u91dd\u5c0d GPT-J \u9032\u884c\u7684\u63a8\u8ad6\u57fa\u6e96\u6e2c\u8a66\u3002GPT-J \u662f\u4e00\u500b\u5177\u6709 60 \u5104\u53c3\u6578\u7684\u5927\u578b\u8a9e\u8a00\u6a21\u578b\uff0c\u800c\u53c3\u6578\u662f\u7528\u4f86\u8861\u91cf\u4eba\u5de5\u667a\u6167\u6a21\u578b\u5927\u5c0f\u7684\u7c97\u7565\u6307\u6a19\u3002<\/p>\n\n\n\n

<\/p>\n\n\n\n

TensorRT-LLM<\/strong>\u5927\u5e45\u63d0\u5347\u63a8\u8ad6\u6548\u80fd<\/strong><\/h3>\n\n\n\n

\u70ba\u4e86\u6e1b\u5c11\u5404\u7a2e\u898f\u6a21\u7684\u8907\u96dc\u5de5\u4f5c\u8ca0\u8f09\uff0cNVIDIA \u958b\u767c\u4e86 TensorRT-LLM<\/a>\uff0c\u9019\u662f\u4e00\u7a2e\u53ef\u6700\u4f73\u5316\u63a8\u8ad6\u7684\u751f\u6210\u5f0f\u4eba\u5de5\u667a\u6167\u8edf\u9ad4\u3002\u9019\u500b\u958b\u6e90\u7a0b\u5f0f\u78bc\u5728\u516b\u6708\u5411 MLPerf \u63d0\u4ea4\u6e2c\u8a66\u7d50\u679c\u6642\u5c1a\u672a\u5b8c\u6210\uff0c\u80fd\u4f7f\u5ba2\u6236\u80fd\u5920\u5728\u7121\u984d\u5916\u6210\u672c\u7684\u60c5\u6cc1\u4e0b\uff0c\u5c07\u5176\u5df2\u8cfc\u8cb7\u7684 H100 GPU \u7684\u63a8\u8ad6\u6548\u80fd\u63d0\u9ad8\u4e00\u500d\u4ee5\u4e0a\u3002<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

<\/p>\n\n\n\n

NVIDIA \u5167\u90e8\u6e2c\u8a66\u986f\u793a\uff0c\u5728 H100 GPU \u4e0a\u4f7f\u7528 TensorRT-LLM\uff0c\u8207\u4ee5\u524d\u7684 GPU \u904b\u884c GPT-J 6B \u76f8\u6bd4\uff0c\u6548\u80fd\u63d0\u5347\u9ad8\u9054 8 \u500d\u3002<\/p>\n\n\n\n

\u9019\u500b\u8edf\u9ad4\u6e90\u65bc NVIDIA \u8207\u696d\u754c\u9818\u5148\u516c\u53f8\u7684\u5408\u4f5c\uff0c\u5305\u62ec Meta\u3001AnyScale\u3001Cohere\u3001Deci\u3001Grammarly\u3001Mistral AI\u3001MosaicML\uff08\u73fe\u70ba Databricks \u7684\u4e00\u90e8\u5206\uff09\u3001OctoML\u3001Tabnine \u548c Together AI\uff0c\u4ee5\u52a0\u901f\u548c\u6700\u4f73\u5316\u5927\u578b\u8a9e\u8a00\u6a21\u578b\u63a8\u8ad6\u7684\u904e\u7a0b\u3002<\/p>\n\n\n\n

MosaicML\u5728TensorRT-LLM \u7684\u57fa\u790e\u4e0a\u589e\u52a0\u6240\u9700\u7684\u529f\u80fd\uff0c\u4e26\u5c07\u5176\u7d0d\u5165\u73fe\u6709\u7684\u670d\u52d9\u5806\u758a\u3002Databricks \u5de5\u7a0b\u90e8\u9580\u526f\u7e3d\u88c1 Naveen Rao \u6307\u51fa\uff1a\u300c\u9019\u7d55\u5c0d\u662f\u4e00\u4ef6\u8f15\u800c\u6613\u8209\u7684\u4e8b\u3002\u300d<\/p>\n\n\n\n

\u300cTensorRT-LLM \u7c21\u55ae\u6613\u7528\u3001\u529f\u80fd\u591a\u6a23\u4e14\u76f8\u7576\u6709\u6548\u7387\u3002\u5b83\u70ba\u4f7f\u7528 NVIDIA GPU \u7684\u5927\u578b\u8a9e\u8a00\u6a21\u578b\u670d\u52d9\u63d0\u4f9b\u4e86\u6700\u5148\u9032\u7684\u6548\u80fd\uff0c\u8b93\u6211\u5011\u80fd\u5920\u628a\u7701\u4e0b\u4f86\u7684\u6210\u672c\u56de\u994b\u7d66\u5ba2\u6236\u3002\u300dRao \u8aaa\u3002<\/p>\n\n\n\n

TensorRT-LLM \u662f NVIDIA \u5168\u7aef AI \u5e73\u53f0\u6301\u7e8c\u4e0d\u65b7\u5275\u65b0\u7684\u6700\u65b0\u7bc4\u4f8b\u3002\u9019\u4e9b\u4e0d\u65b7\u5347\u7d1a\u7684\u8edf\u9ad4\u70ba\u7528\u6236\u63d0\u4f9b\u4e86\u53ef\u96a8\u6642\u9593\u589e\u9577\u7684\u6027\u80fd\uff0c\u800c\u7121\u9700\u984d\u5916\u6210\u672c\uff0c\u4e26\u4e14\u80fd\u9069\u61c9\u7576\u4eca\u591a\u6a23\u5316\u7684\u4eba\u5de5\u667a\u6167\u5de5\u4f5c\u8ca0\u8f09\u3002<\/p>\n\n\n\n

<\/p>\n\n\n\n

L4<\/strong>\u63d0\u5347\u4e3b\u6d41\u4f3a\u670d\u5668\u7684\u63a8\u8ad6\u6548\u80fd<\/strong><\/h3>\n\n\n\n

\u5728\u6700\u65b0\u7684 MLPerf \u57fa\u6e96\u6e2c\u8a66\u4e2d\uff0cNVIDIA L4 GPU<\/a> \u5728\u5404\u7a2e\u5de5\u4f5c\u8ca0\u8f09\u4e0a\u8868\u73fe\u5353\u8d8a\uff0c\u63d0\u4f9b\u5168\u9762\u6027\u7684\u51fa\u8272\u6027\u80fd\u3002<\/p>\n\n\n\n

\u4f8b\u5982\uff0cL4 GPU \u904b\u884c\u5728\u7cbe\u5de7\u3001\u529f\u8017\u70ba 72W \u7684\u8f49\u63a5\u5361\u4e0a\uff0c\u8207\u529f\u8017\u9ad8\u51fa\u8fd1 5 \u500d\u7684 CPU \u76f8\u6bd4\u8f03\uff0cL4 GPU \u63d0\u4f9b\u9ad8\u51fa 6 \u500d\u6548\u80fd\u3002<\/p>\n\n\n\n

\u9664\u6b64\u4e4b\u5916\uff0cL4 GPU \u5167\u5efa\u5c08\u5c6c\u7684\u5a92\u9ad4\u5f15\u64ce\uff0c\u5728 NVIDIA \u7684\u6e2c\u8a66\u4e2d\u8207 CUDA \u8edf\u9ad4\u5408\u7528\u80fd\u52a0\u901f\u96fb\u8166\u8996\u89ba\u61c9\u7528\u9054 120 \u500d\u3002<\/p>\n\n\n\n

\u76ee\u524d\u53ef\u4ee5\u5f9e Google Cloud \u548c\u8a31\u591a\u7cfb\u7d71\u88fd\u9020\u5546\u7aef\u4f7f\u7528 L4 GPU\u3002\u5b83\u5011\u70ba\u5f9e\u6d88\u8cbb\u8005\u7db2\u8def\u670d\u52d9\u5230\u85e5\u7269\u7814\u767c\u7b49\u591a\u500b\u7522\u696d\u7684\u5ba2\u6236\u63d0\u4f9b\u670d\u52d9\u3002<\/p>\n\n\n\n

<\/p>\n\n\n\n

\u5728\u908a\u7de3\u74b0\u5883\u4e2d\u6548\u80fd\u63d0\u5347<\/strong><\/strong><\/h3>\n\n\n\n

\u6b64\u5916\uff0cNVIDIA \u61c9\u7528\u4e86\u65b0\u7684\u6a21\u578b\u58d3\u7e2e\u6280\u8853\uff0c\u4f7f\u5728 L4 GPU \u4e0a\u904b\u884c BERT LLM \u7684\u6548\u80fd\u63d0\u5347\u9054 4.7 \u500d\u3002\u9019\u4e00\u7d50\u679c\u5728 MLPerf \u7684\u6240\u8b02\u958b\u653e\u7d44\u5225\uff08Open Division\uff09\u4e2d\u5be6\u73fe\uff0c\u9019\u662f\u7528\u65bc\u5c55\u793a\u65b0\u80fd\u529b\u7684\u4e00\u500b\u985e\u5225\u3002<\/p>\n\n\n\n

\u8a72\u6280\u8853\u9810\u8a08\u5c07\u9069\u7528\u65bc\u6240\u6709\u4eba\u5de5\u667a\u6167\u5de5\u4f5c\u8ca0\u8f09\u3002\u7576\u5728\u5c3a\u5bf8\u548c\u529f\u8017\u53d7\u9650\u7684\u908a\u7de3\u8a2d\u5099\u4e0a\u904b\u884c\u6a21\u578b\u6642\uff0c\u5b83\u5c24\u5176\u6709\u50f9\u503c\u3002<\/p>\n\n\n\n

\u5728\u53e6\u4e00\u500b\u908a\u7de3\u904b\u7b97\u9818\u5148\u7bc4\u4f8b\u4e2d\uff0cNVIDIA Jetson Orin<\/a> \u7cfb\u7d71\u6a21\u7d44\u986f\u793a\u7269\u4ef6\u5075\u6e2c\u7684\u6548\u80fd\u76f8\u5c0d\u524d\u4e00\u8f2a\u6e2c\u8a66\u63d0\u5347\u9ad8\u9054 84%\uff0c\u9019\u662f\u908a\u7de3\u4eba\u5de5\u667a\u6167\u548c\u6a5f\u5668\u4eba\u5834\u666f\u4e2d\u5e38\u898b\u7684\u96fb\u8166\u8996\u89ba\u4f7f\u7528\u6848\u4f8b\u3002<\/p>\n\n\n

\n
\"\"<\/figure><\/div>\n\n\n

<\/p>\n\n\n\n

Jetson Orin \u7684\u5148\u884c\u7522\u54c1\u4f86\u81ea\u63a1\u7528\u6700\u65b0\u7248\u6676\u7247\u6838\u5fc3\u7684\u8edf\u9ad4\uff0c\u5982\u53ef\u7a0b\u5f0f\u8a2d\u8a08\u8996\u89ba\u52a0\u901f\u5668\u3001NVIDIA Ampere\u67b6\u69cb<\/a>GPU\u548c\u5c08\u7528\u6df1\u5ea6\u5b78\u7fd2\u52a0\u901f\u5668\u3002<\/p>\n\n\n\n

\u591a\u529f\u80fd\u7684\u6548\u80fd\uff0c\u5ee3\u5927\u7684\u751f\u614b\u7cfb\u7d71<\/strong><\/strong><\/p>\n\n\n\n

MLPerf \u57fa\u6e96\u662f\u900f\u660e\u4e14\u5ba2\u89c0\u7684\uff0c\u56e0\u6b64\u4f7f\u7528\u8005\u53ef\u4ee5\u4f9d\u9760\u5176\u7d50\u679c\u505a\u51fa\u660e\u667a\u7684\u8cfc\u8cb7\u6c7a\u7b56\u3002\u5b83\u5011\u6db5\u84cb\u4e86\u5ee3\u6cdb\u7684\u61c9\u7528\u6848\u4f8b\u548c\u60c5\u666f\uff0c\u56e0\u6b64\u4f7f\u7528\u8005\u77e5\u9053\u4ed6\u5011\u53ef\u4ee5\u7372\u5f97\u53ef\u9760\u4e14\u90e8\u7f72\u9748\u6d3b\u7684\u6548\u80fd\u3002<\/p>\n\n\n\n

\u5728\u672c\u8f2a\u6e2c\u8a66\u4e2d\u53c3\u8207\u63d0\u4ea4\u7684\u5408\u4f5c\u5925\u4f34\u5305\u62ec\u96f2\u7aef\u670d\u52d9\u4f9b\u61c9\u5546 Microsoft Azure \u548c Oracle Cloud Infrastructure\uff0c\u4ee5\u53ca\u83ef\u78a9\u96fb\u8166\u3001Connect Tech\u3001\u6234\u723e\u79d1\u6280\u96c6\u5718\u3001\u5bcc\u58eb\u901a\u516c\u53f8\u3001\u6280\u5609\u79d1\u6280\u3001\u6167\u8207\u79d1\u6280\u3001\u806f\u60f3\u96c6\u5718\u3001\u96f2\u9054\u79d1\u6280\u548c\u7f8e\u8d85\u5fae\u7b49\u7cfb\u7d71\u88fd\u9020\u5546\u3002<\/p>\n\n\n\n

\u7e3d\u9ad4\u4f86\u8aaa\uff0cMLPerf \u5f97\u5230\u4e86\u8d85\u904e 70 \u5bb6\u7d44\u7e54\u7684\u652f\u6301\uff0c\u5305\u62ec\u963f\u91cc\u5df4\u5df4\u3001Arm\u3001\u601d\u79d1\u3001Google\u3001\u54c8\u4f5b\u5927\u5b78\u3001\u82f1\u7279\u723e\u3001Meta\u3001\u5fae\u8edf\u548c\u591a\u502b\u591a\u5927\u5b78\u7b49\u3002<\/p>\n\n\n\n

\u6b32\u77ad\u89e3\u66f4\u591a\u8a73\u7d30\u8cc7\u8a0a\u4ee5\u53ca\u6211\u5011\u5982\u4f55\u7372\u5f97\u9019\u4e9b\u6210\u679c\uff0c\u8acb\u95b1\u8b80\u6280\u8853\u90e8\u843d\u683c\u6587\u7ae0<\/a>\u3002<\/p>\n\n\n\n

\u65bc\u6b64\u6b21\u6e2c\u8a66\u4e2d\u4f7f\u7528\u7684\u5404\u7a2e\u8edf\u9ad4\u516c\u958b\u65bc MLPerf \u8cc7\u6e90\u5eab\uff0c\u6bcf\u500b\u4eba\u90fd\u80fd\u53d6\u5f97\u9019\u4e9b\u4e16\u754c\u7d1a\u7684\u6210\u679c\u3002\u6211\u5011\u4e0d\u65b7\u5c07\u6700\u4f73\u5316\u7d50\u679c\u653e\u5165 NVIDIA NGC<\/a> \uff08GPU \u52a0\u901f\u8edf\u9ad4\u76ee\u9304\uff09\u7684\u5bb9\u5668\u4e2d\uff0c\u63d0\u4f9b GPU \u61c9\u7528\u3002<\/p>\n\n\n\n

<\/p>\n\n\n\n

<\/p>\n\n\n\n

<\/p>\n\n\n\n

\ud83d\udfe6ioioTIMES \u8981\u9001\u60a8SSD \u5feb\u4f86\u5b8c\u6210\u4efb\u52d9\u8a66\u624b\u6c23<\/a>
\ud83d\udfe6\u73fe\u5728\u5c31\u52a0\u5165 ioioTIMES \u81c9\u66f8\u7c89\u7d72\u5718<\/a> \u66f4\u591a\u4e92\u52d5\u3001\u66f4\u591a\u597d\u5eb7\u650f\u62b5\u52a0!!<\/strong>
\ud83d\udfe6\u6211\u5011\u6709LINE TODAY<\/a>\u983b\u9053\u4e86\uff0c\u5feb\u4f86\u8ffd\u8e2a\u6211\u5011\u5427!!–\u6700\u65b0\u79d1\u6280\u65b0\u805e \u76e1\u5728\u4f60\u624b<\/strong><\/h4>\n\n\n\n

<\/p>\n","protected":false},"excerpt":{"rendered":"

NVIDIA GH200\u3001H100\u548c<\/p>\n","protected":false},"author":3,"featured_media":43499,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"rank_math_lock_modified_date":false,"footnotes":""},"categories":[13],"tags":[5302,6360,71,7184,7183],"class_list":["post-43497","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-news","tag-grace-hopper","tag-mlperf","tag-nvidia","tag-7184","tag-7183"],"_links":{"self":[{"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/posts\/43497"}],"collection":[{"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=43497"}],"version-history":[{"count":3,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/posts\/43497\/revisions"}],"predecessor-version":[{"id":43503,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/posts\/43497\/revisions\/43503"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/media\/43499"}],"wp:attachment":[{"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=43497"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=43497"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=43497"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}