{"id":49941,"date":"2023-11-16T16:27:00","date_gmt":"2023-11-16T08:27:00","guid":{"rendered":"https:\/\/www.ioiotimes.com\/?p=49941"},"modified":"2023-11-17T10:32:28","modified_gmt":"2023-11-17T02:32:28","slug":"%e9%bb%9e%e4%ba%ae%e6%9c%aa%e4%be%86%ef%bc%9atensorrt-llm%e7%99%bc%e5%b8%83%e5%8a%a0%e9%80%9fai%e6%8e%a8%e8%ab%96%e6%95%88%e8%83%bd%ef%bc%8c%e6%94%af%e6%8f%b4%e5%9c%a8rtx%e9%a9%85%e5%8b%95%e7%9a%84win","status":"publish","type":"post","link":"https:\/\/www.ioiotimes.com\/?p=49941","title":{"rendered":"\u9ede\u4eae\u672a\u4f86\uff1aTensorRT-LLM\u767c\u5e03\u52a0\u901fAI\u63a8\u8ad6\u6548\u80fd\uff0c\u652f\u63f4\u5728RTX\u9a45\u52d5\u7684Windows PC\u4e0a\u904b\u884c\u7684\u65b0\u6a21\u578b"},"content":{"rendered":"\n<h3 class=\"wp-block-heading\"><strong>\u65bc\u5fae\u8edfIgnite\u5168\u7403\u6280\u8853\u5927\u6703\u767c\u5e03\u7684\u65b0\u5de5\u5177\u548c\u8cc7\u6e90\u5305\u62ecOpenAI Chat API\u7684TensorRT-LLM\u5c01\u88dd\u5668\u3001DirectML\u70baLlama 2\u5e36\u4f86RTX\u9a45\u52d5\u7684\u6548\u80fd\u63d0\u5347\u3001\u5176\u4ed6\u71b1\u9580LLM<\/strong><\/h3>\n\n\n\n<p>Windows PC \u4e0a\u7684\u4eba\u5de5\u667a\u6167\uff08AI\uff09\u6a19\u8a8c\u8457\u79d1\u6280\u53f2\u4e0a\u7684\u95dc\u9375\u6642\u523b\uff0c\u5b83\u5c07\u5fb9\u5e95\u6539\u8b8a\u73a9\u5bb6\u3001\u5275\u4f5c\u8005\u3001\u76f4\u64ad\u4e3b\u3001\u4e0a\u73ed\u65cf\u3001\u5b78\u751f\u4e43\u81f3\u666e\u901a PC \u7528\u6236\u7684\u9ad4\u9a57\u3002<\/p>\n\n\n\n<p>AI \u70ba 1 \u5104\u591a\u53f0\u63a1\u7528 RTX GPU \u7684 Windows PC \u548c\u5de5\u4f5c\u7ad9\u63d0\u9ad8\u751f\u7522\u529b\uff0c\u5e36\u4f86\u524d\u6240\u672a\u6709\u7684\u6a5f\u6703\u3002NVIDIA RTX \u6280\u8853\u4f7f\u958b\u767c\u8005\u66f4\u8f15\u9b06\u5730\u5275\u5efa AI \u61c9\u7528\uff0c\u5f9e\u800c\u6539\u8b8a\u4eba\u5011\u4f7f\u7528\u96fb\u8166\u7684\u65b9\u5f0f\u3002<\/p>\n\n\n\n<p>\u5728\u5fae\u8edf Ignite \u5927\u6703\u4e0a\u767c\u5e03\u7684\u5168\u65b0\u6700\u4f73\u5316\u3001\u6a21\u578b\u548c\u8cc7\u6e90\uff0c\u5c07\u8b93\u958b\u767c\u8005\u53ef\u4ee5\u66f4\u5feb\u63d0\u4f9b\u65b0\u7684\u7d42\u7aef\u4f7f\u7528\u8005\u9ad4\u9a57\u3002<\/p>\n\n\n\n<p><a href=\"https:\/\/developer.nvidia.com\/tensorrt\" target=\"_blank\" rel=\"noreferrer noopener\">TensorRT-LLM<\/a> \u662f\u4e00\u6b3e\u63d0\u5347 AI \u63a8\u8ad6\u6548\u80fd\u7684\u958b\u6e90\u8edf\u9ad4\uff0c\u5b83\u5373\u5c07\u767c\u5e03\u7684\u66f4\u65b0\u7248\u5c07\u652f\u6301\u66f4\u591a<a href=\"https:\/\/www.nvidia.com\/en-us\/glossary\/data-science\/large-language-models\/\" target=\"_blank\" rel=\"noreferrer noopener\">\u5927\u578b\u8a9e\u8a00\u6a21\u578b<\/a>\uff0c\u53ef\u5728 RTX GPU 8GB VRAM \u4ee5\u4e0a\u7684 PC \u548c\u7b46\u8a18\u578b\u96fb\u8166\u4e0a\uff0c\u4f7f\u8981\u6c42\u56b4\u82db\u7684 AI \u5de5\u4f5c\u8ca0\u8f09\u66f4\u5bb9\u6613\u5b8c\u6210\u3002<\/p>\n\n\n\n<p>Tensor RT-LLM for Windows \u5373\u5c07\u900f\u904e\u5168\u65b0\u5c01\u88dd\u4ecb\u9762\u8207 OpenAI \u5ee3\u53d7\u6b61\u8fce\u7684\u804a\u5929 API \u76f8\u5bb9\u3002\u9019\u5c07\u4f7f\u6578\u4ee5\u767e\u8a08\u7684\u958b\u767c\u8005\u5c08\u6848\u548c\u61c9\u7528\u80fd\u5728 RTX PC \u7684\u672c\u5730\u904b\u884c\uff0c\u800c\u975e\u96f2\u7aef\u904b\u884c\uff0c\u56e0\u6b64\u7528\u6236\u53ef\u4ee5\u5728 PC \u4e0a\u4fdd\u7559\u79c1\u4eba\u548c\u5c08\u6709\u8cc7\u6599\u3002<\/p>\n\n\n\n<p>\u5ba2\u88fd\u7684\u751f\u6210\u5f0f AI \u9700\u8981\u6642\u9593\u548c\u7cbe\u529b\u4f86\u7dad\u8b77\u5c08\u6848\u3002\u7279\u5225\u662f\u8de8\u591a\u500b\u74b0\u5883\u548c\u5e73\u81fa\u9032\u884c\u5354\u4f5c\u548c\u90e8\u7f72\u6642\uff0c\u8a72\u904e\u7a0b\u53ef\u80fd\u6703\u7570\u5e38\u8907\u96dc\u548c\u8017\u6642\u3002<\/p>\n\n\n\n<p><a href=\"https:\/\/www.nvidia.com\/zh-tw\/deep-learning-ai\/solutions\/data-science\/workbench\/\" target=\"_blank\" rel=\"noreferrer noopener\">NVIDIA \u8cc7\u6599\u79d1\u5b78\u5de5\u4f5c\u53f0\uff08AI Workbench\uff09<\/a> \u662f\u4e00\u500b\u7d71\u4e00\u3001\u6613\u7528\u7684\u5de5\u5177\u5305\uff0c\u5141\u8a31\u958b\u767c\u8005\u5728 PC \u6216\u5de5\u4f5c\u7ad9\u4e0a\u5feb\u901f\u5efa\u7acb\u3001\u6e2c\u8a66\u548c\u5ba2\u88fd\u9810\u5148\u8a13\u7df4\u7684\u751f\u6210\u5f0f AI \u6a21\u578b\u548c LLM\u3002\u5b83\u70ba\u958b\u767c\u8005\u63d0\u4f9b\u4e00\u500b\u55ae\u4e00\u5e73\u81fa\uff0c\u7528\u65bc\u7d44\u7e54\u4ed6\u5011\u7684 AI \u5c08\u6848\uff0c\u4e26\u6839\u64da\u7279\u5b9a\u4f7f\u7528\u8005\u9700\u6c42\u4f86\u8abf\u6574\u6a21\u578b\u3002<\/p>\n\n\n\n<p>\u9019\u4f7f\u958b\u767c\u8005\u80fd\u5920\u9032\u884c\u7121\u7e2b\u5354\u4f5c\u548c\u90e8\u7f72\uff0c\u5feb\u901f\u5275\u5efa\u5177\u6709\u6210\u672c\u6548\u76ca\u3001\u53ef\u64f4\u5c55\u7684\u751f\u6210\u5f0f AI \u6a21\u578b\u3002<a href=\"https:\/\/developer.nvidia.com\/ai-workbench-early-access\" target=\"_blank\" rel=\"noreferrer noopener\">\u6b61\u8fce\u52a0\u5165\u6436\u5148\u9ad4\u9a57\u540d\u55ae<\/a>\uff0c\u6210\u70ba\u9996\u6279\u7528\u6236\u4ee5\u7387\u5148\u77ad\u89e3\u4e0d\u65b7\u66f4\u65b0\u7684\u529f\u80fd\uff0c\u4e26\u63a5\u6536\u66f4\u65b0\u8cc7\u8a0a\u3002<\/p>\n\n\n\n<p>\u70ba\u652f\u63f4 AI \u958b\u767c\u8005\uff0cNVIDIA \u8207\u5fae\u8edf\u5171\u540c\u767c\u5e03 DirectML \u589e\u5f37\u529f\u80fd\u4ee5\u52a0\u901f Llama 2\uff0c\u6642\u4e0b\u6700\u71b1\u9580\u7684\u57fa\u790e AI \u6a21\u578b\u4e4b\u4e00\u3002\u9664\u4e86<a href=\"https:\/\/blogs.nvidia.com\/blog\/2023\/10\/17\/tensorrt-llm-windows-stable-diffusion-rtx\/\" target=\"_blank\" rel=\"noreferrer noopener\">\u5236\u5b9a\u4e00\u500b\u5168\u65b0\u7684\u6548\u80fd\u6a19\u6e96<\/a>\uff0c\u958b\u767c\u8005\u73fe\u5728\u4ea6\u6709\u66f4\u591a\u8de8\u4f9b\u61c9\u5546\u90e8\u7f72\u53ef\u9078\u64c7\u3002<\/p>\n\n\n\n<p><\/p>\n\n\n\n<p><strong>\u53ef\u651c\u5f0f<\/strong><strong>AI<\/strong><\/p>\n\n\n\n<p>NVIDIA \u65bc 10 \u6708\u767c\u5e03\u4e86\u7528\u65bc\u52a0\u901f\u5927\u578b\u8a9e\u8a00\u6a21\u578b\uff08LLM\uff09\u63a8\u7406\u7684\u51fd\u5f0f\u5eab TensorRT-LLM for Windows\u3002\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0<\/p>\n\n\n\n<p>\u672c\u6708\u5e95\u767c\u5e03\u7684 TensorRT-LLM v0.6.0 \u66f4\u65b0\u5c07\u5e36\u4f86\u9ad8\u9054 5 \u500d\u7684\u63a8\u8ad6\u6548\u80fd\u63d0\u5347\uff0c\u4e26\u652f\u63f4\u66f4\u591a\u71b1\u9580\u7684 LLM\uff0c\u5305\u62ec\u5168\u65b0 Mistral 7B \u548c Nemotron-3 8B\u3002\u9019\u4e9b LLM \u7248\u672c\u5c07\u53ef\u5728\u6240\u6709\u63a1\u7528 8GB \u8a18\u61b6\u9ad4\u4ee5\u4e0a\u7684 GeForce RTX 30 \u7cfb\u5217\u548c 40 \u7cfb\u5217 GPU \u4e0a\u904b\u884c\uff0c\u5f9e\u800c\u4f7f\u6700\u8f15\u4fbf\u7684 Windows PC \u8a2d\u5099\u4e5f\u80fd\u5177\u5099\u65bc\u672c\u5730\u5feb\u901f\u4e14\u6e96\u78ba\u904b\u884c LLM \u7684\u529f\u80fd\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"1660\" height=\"1635\" src=\"https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2023\/11\/20231116-nvidia07.png\" alt=\"\" class=\"wp-image-49961\" title=\"\" srcset=\"https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2023\/11\/20231116-nvidia07.png 1660w, https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2023\/11\/20231116-nvidia07-300x295.png 300w, https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2023\/11\/20231116-nvidia07-1024x1009.png 1024w, https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2023\/11\/20231116-nvidia07-768x756.png 768w, https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2023\/11\/20231116-nvidia07-1536x1513.png 1536w\" sizes=\"(max-width: 1660px) 100vw, 1660px\" \/><figcaption class=\"wp-element-caption\">\u25b2TensorRT-LLM v0.6.0 \u7248\u5e36\u4f86\u9ad8\u90545 \u500d\u7684\u63a8\u8ad6\u6548\u80fd\u63d0\u5347<\/figcaption><\/figure>\n\n\n\n<p><\/p>\n\n\n\n<p><\/p>\n\n\n\n<p>\u65b0\u767c\u5e03\u7684 TensorRT-LLM \u53ef\u5728 <a href=\"https:\/\/developer.nvidia.com\/blog\/optimizing-inference-on-llms-with-tensorrt-llm-now-publicly-available\/\" target=\"_blank\" rel=\"noreferrer noopener\">\/NVIDIA\/TensorRT-LLM<\/a> GitHub \u5132\u5b58\u5eab\u4e2d\u4e0b\u8f09\u5b89\u88dd\uff0c\u5168\u65b0\u6700\u4f73\u5316\u7684\u6a21\u578b\u5c07\u63d0\u4f9b\u65bc <a href=\"http:\/\/ngc.nvidia.com\/\" target=\"_blank\" rel=\"noreferrer noopener\">ngc.nvidia.com<\/a>\u3002<\/p>\n\n\n\n<p><\/p>\n\n\n\n<p><strong>\u5be6\u73fe\u5f9e\u5bb9\u5c0d\u8a71<\/strong><strong><\/strong><\/p>\n\n\n\n<p>\u4e16\u754c\u5404\u5730\u7684\u958b\u767c\u8005\u548c\u611b\u597d\u8005\u5c07 OpenAI \u7684\u804a\u5929 API \u5ee3\u6cdb\u7528\u65bc\u5404\u7a2e\u61c9\u7528\uff0c\u5305\u62ec\u7e3d\u7d50\u7db2\u9801\u5167\u5bb9\u3001\u8d77\u64ec\u6587\u4ef6\u8349\u7a3f\u548c\u96fb\u5b50\u90f5\u4ef6\u3001\u5206\u6790\u4e26\u8996\u89ba\u5316\u8cc7\u6599\u3001\u88fd\u4f5c\u7c21\u5831\u7b49\u3002<\/p>\n\n\n\n<p>\u9019\u985e\u4ee5\u96f2\u7aef\u70ba\u57fa\u790e\u7684 AI \u9762\u81e8\u7684\u4e00\u5927\u6311\u6230\u5728\u65bc\u5b83\u5011\u9700\u8981\u4f7f\u7528\u8005\u4e0a\u50b3\u8f38\u5165\u8cc7\u6599\uff0c\u56e0\u6b64\u5c0d\u65bc\u79c1\u4eba\u6216\u5c08\u6709\u8cc7\u6599\u6216\u8655\u7406\u5927\u578b\u8cc7\u6599\u96c6\u800c\u8a00\u4e26\u4e0d\u5be6\u7528\u3002<\/p>\n\n\n\n<p>\u70ba\u61c9\u5c0d\u9019\u4e00\u6311\u6230\uff0cNVIDIA \u5373\u5c07\u555f\u7528 TensorRT-LLM for Windows\uff0c\u85c9\u7531\u5168\u65b0\u5c01\u88dd\u5668\u63d0\u4f9b\u8207 OpenAI \u5ee3\u53d7\u6b61\u8fce\u7684 ChatAPI \u985e\u4f3c\u7684 API \u4ecb\u9762\uff0c\u70ba\u958b\u767c\u8005\u5e36\u4f86\u985e\u4f3c\u7684\u5de5\u4f5c\u6d41\u7a0b\uff0c\u7121\u8ad6\u4ed6\u5011\u8a2d\u8a08\u7684\u6a21\u578b\u548c\u61c9\u7528\u8981\u5728 RTX PC \u7684\u672c\u5730\u904b\u884c\u3001\u5728\u96f2\u7aef\u904b\u884c\uff0c\u53ea\u9700\u4fee\u6539\u4e00\u5230\u5169\u884c\u4ee3\u78bc\uff0c\u6578\u767e\u500bAI\u9a45\u52d5\u7684\u958b\u767c\u8005\u5c08\u6848\u548c\u61c9\u7528\u73fe\u5728\u5c31\u80fd\u5f9e\u5feb\u901f\u7684\u672c\u5730AI\u4e2d\u53d7\u76ca\u3002\u4f7f\u7528\u8005\u53ef\u5c07\u8cc7\u6599\u4fdd\u5b58\u5728 PC \u4e0a\uff0c\u4e0d\u5fc5\u64d4\u5fc3\u5c07\u8cc7\u6599\u4e0a\u50b3\u5230\u96f2\u7aef\u3002<\/p>\n\n\n\n<p>\u6b64\u5916\uff0c\u6700\u91cd\u8981\u7684\u4e00\u9ede\u662f\u9019\u4e9b\u9805\u76ee\u548c\u61c9\u7528\u4e2d\u6709\u5f88\u591a\u90fd\u662f\u958b\u6e90\u7684\uff0c\u958b\u767c\u8005\u53ef\u4ee5\u8f15\u9b06\u5229\u7528\u548c\u64f4\u5c55\u5b83\u5011\u7684\u529f\u80fd\uff0c\u5f9e\u800c\u52a0\u901f\u751f\u6210\u5f0f AI \u5728 RTX \u9a45\u52d5\u7684 Windows PC \u4e0a\u7684\u61c9\u7528\u3002<\/p>\n\n\n\n<p>\u8a72\u5c01\u88dd\u5668\u53ef\u8207\u6240\u6709\u5c0d TensorRT-LLM \u9032\u884c\u6700\u4f73\u5316\u7684 LLM\uff08\u5982\uff1aLlama 2\u3001Mistral \u548c NV LLM\uff09\u914d\u5408\u4f7f\u7528\uff0c\u4e26\u4f5c\u70ba\u53c3\u8003\u5c08\u6848\u5728 GitHub \u4e0a\u767c\u5e03\uff0c\u540c\u6b65\u4e5f\u5c07\u767c\u5e03\u53ef\u4ee5\u5728 RTX \u4e0a\u4f7f\u7528 LLM \u7684\u5176\u4ed6\u958b\u767c\u8005\u8cc7\u6e90\u3002<\/p>\n\n\n\n<p><\/p>\n\n\n\n<p><strong>\u6a21\u578b\u52a0\u901f<\/strong><strong><\/strong><\/p>\n\n\n\n<p>\u958b\u767c\u8005\u73fe\u53ef\u5229\u7528\u5c16\u7aef\u7684 AI \u6a21\u578b\uff0c\u4e26\u900f\u904e\u8de8\u4f9b\u61c9\u5546 API \u9032\u884c\u90e8\u7f72\u3002NVIDIA \u548c\u5fae\u8edf\u4e00\u76f4\u81f4\u529b\u65bc\u589e\u5f37\u958b\u767c\u8005\u80fd\u529b\uff0c\u900f\u904e DirectML API \u5728 RTX \u4e0a\u52a0\u901f Llama\u3002<\/p>\n\n\n\n<p>\u5728 10 \u6708\u5df2\u5ba3\u5e03\uff0c\u5728\u70ba\u9019\u4e9b\u6a21\u578b\u63d0\u4f9b\u6700\u5feb\u63a8\u8ad6\u6548\u80fd\u7684\u57fa\u790e\u4e0a\uff0c\u6b64\u9805\u8de8\u4f9b\u61c9\u5546\u90e8\u7f72\u7684\u5168\u65b0\u9078\u9805\u5c07\u4f7f AI \u529f\u80fd\u5c0e\u5165 PC \u8b8a\u5f97\u524d\u6240\u672a\u6709\u7684\u7c21\u55ae\u3002<\/p>\n\n\n\n<p>\u958b\u767c\u8005\u548c\u611b\u597d\u8005\u53ef\u4e0b\u8f09\u6700\u65b0\u7684 ONNX \u57f7\u884c\u6642\u9593\u3001\u4f9d\u7167\u5fae\u8edf\u7684\u5b89\u88dd\u8aaa\u660e\u9032\u884c\u64cd\u4f5c\uff0c\u4e26\u5b89\u88dd<a href=\"https:\/\/www.nvidia.com\/download\/index.aspx\" target=\"_blank\" rel=\"noreferrer noopener\">\u6700\u65b0NVIDIA\u9a45\u52d5\u7a0b\u5f0f<\/a>\uff08\u5c07\u65bc 11 \u6708 21 \u65e5\u767c\u4f48\uff09\uff0c\u4ee5\u7372\u5f97\u6700\u65b0\u7684\u6700\u4f73\u5316\u9ad4\u9a57\u3002<\/p>\n\n\n\n<p>\u9019\u4e9b\u5168\u65b0\u7684\u6700\u4f73\u5316\u3001\u6a21\u578b\u548c\u8cc7\u6e90\u5c07\u52a0\u901f AI \u529f\u80fd\u548c\u61c9\u7528\u5728\u5168\u7403 1 \u5104\u53f0 RTX PC \u4e0a\u7684\u958b\u767c\u548c\u90e8\u7f72\uff0c\u540c\u6642\u52a0\u5165\u4e86 400 \u591a\u500b\u5408\u4f5c\u5925\u4f34\u7684\u884c\u5217\uff0c\u9019\u4e9b\u5925\u4f34\u7686\u5df2\u63a8\u51fa\u7531 RTX GPU \u52a0\u901f\u7684 AI \u9a45\u52d5\u7684\u61c9\u7528\u548c\u904a\u6232\u3002<\/p>\n\n\n\n<p>\u96a8\u8457\u6a21\u578b\u6613\u7528\u6027\u7684\u63d0\u9ad8\uff0c\u800c\u958b\u767c\u8005\u5c07\u66f4\u591a\u751f\u6210\u5f0f AI \u529f\u80fd\u5e36\u5230 RTX \u9a45\u52d5\u7684 Windows PC \u4e0a\uff0cRTX GPU \u5c07\u6210\u70ba\u7528\u6236\u63a1\u7528\u9019\u4e00\u5f37\u5927\u6280\u8853\u7684\u95dc\u9375\u3002<\/p>\n\n\n\n<p><\/p>\n\n\n\n<p><\/p>\n\n\n\n<h4 class=\"wp-block-heading has-text-align-right has-very-light-gray-to-cyan-bluish-gray-gradient-background has-background\">\ud83d\udfe6<strong>\u73fe\u5728\u5c31\u52a0\u5165&nbsp;<a href=\"https:\/\/www.facebook.com\/profile.php?id=100086628162118\" target=\"_blank\" rel=\"noreferrer noopener\">ioioTIMES \u81c9\u66f8\u7c89\u7d72\u5718<\/a>&nbsp;\u66f4\u591a\u4e92\u52d5\u3001\u66f4\u591a\u597d\u5eb7\u650f\u62b5\u52a0!!<\/strong><br>\ud83d\udfe6<strong>\u6211\u5011\u6709<a href=\"https:\/\/today.line.me\/tw\/v2\/publisher\/103117\" target=\"_blank\" rel=\"noreferrer noopener nofollow\">LINE TODAY<\/a>\u983b\u9053\u4e86\uff0c\u5feb\u4f86\u8ffd\u8e2a\u6211\u5011\u5427!!&#8211;\u6700\u65b0\u79d1\u6280\u65b0\u805e \u76e1\u5728\u4f60\u624b<\/strong><\/h4>\n","protected":false},"excerpt":{"rendered":"<p>\u65bc\u5fae\u8edfIgnite\u5168\u7403\u6280\u8853\u5927\u6703\u767c\u5e03\u7684<\/p>\n","protected":false},"author":3,"featured_media":49961,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"rank_math_lock_modified_date":false,"footnotes":""},"categories":[13],"tags":[7867,71,7868,7182,7869],"class_list":["post-49941","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-news","tag-ai","tag-nvidia","tag-rtx","tag-tensorrt-llm","tag-ignite"],"_links":{"self":[{"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/posts\/49941"}],"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=49941"}],"version-history":[{"count":4,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/posts\/49941\/revisions"}],"predecessor-version":[{"id":49962,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/posts\/49941\/revisions\/49962"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/media\/49961"}],"wp:attachment":[{"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=49941"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=49941"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=49941"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}