{"id":47999,"date":"2023-10-31T15:30:25","date_gmt":"2023-10-31T07:30:25","guid":{"rendered":"https:\/\/www.ioiotimes.com\/?p=47999"},"modified":"2023-10-31T15:30:26","modified_gmt":"2023-10-31T07:30:26","slug":"%e8%a8%ad%e8%a8%88%e5%b8%ab%e5%88%a9%e7%94%a8%e7%94%9f%e6%88%90%e5%bc%8f%e4%ba%ba%e5%b7%a5%e6%99%ba%e6%85%a7%e4%bd%9c%e7%82%ba%e6%99%b6%e7%89%87%e8%bc%94%e5%8a%a9%ef%bc%8c%e5%8d%8a%e5%b0%8e%e9%ab%94","status":"publish","type":"post","link":"https:\/\/www.ioiotimes.com\/?p=47999","title":{"rendered":"\u8a2d\u8a08\u5e2b\u5229\u7528\u751f\u6210\u5f0f\u4eba\u5de5\u667a\u6167\u4f5c\u70ba\u6676\u7247\u8f14\u52a9\uff0c\u534a\u5c0e\u9ad4\u5de5\u7a0b\u5e2b\u5c55\u793a\u5c08\u696d\u7522\u696d\u5982\u4f55\u4f7f\u7528NVIDIA NeMo\u5ba2\u88fd\u5316\u5927\u578b\u8a9e\u8a00\u6a21\u578b\u4f86\u7372\u5f97\u512a\u52e2"},"content":{"rendered":"<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"600\" height=\"319\" src=\"https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2023\/10\/20231031-nvidia01.jpg\" alt=\"\" class=\"wp-image-48000\" title=\"\" srcset=\"https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2023\/10\/20231031-nvidia01.jpg 600w, https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2023\/10\/20231031-nvidia01-300x160.jpg 300w\" sizes=\"(max-width: 600px) 100vw, 600px\" \/><figcaption class=\"wp-element-caption\">\u25b2\u8a2d\u8a08\u5e2b\u5229\u7528\u751f\u6210\u5f0f\u4eba\u5de5\u667a\u6167\u4f5c\u70ba\u6676\u7247\u8f14\u52a9<\/figcaption><\/figure><\/div>\n\n\n<p><\/p>\n\n\n\n<p>\u4eca\u5929\u767c\u5e03\u7684\u4e00\u7bc7<a href=\"https:\/\/research.nvidia.com\/publication\/2023-10_chipnemo-domain-adapted-llms-chip-design\" target=\"_blank\" rel=\"noreferrer noopener\">\u7814\u7a76\u8ad6\u6587<\/a>\u63cf\u8ff0\u4e86<a href=\"https:\/\/www.nvidia.com\/en-us\/glossary\/data-science\/generative-ai\/\" target=\"_blank\" rel=\"noreferrer noopener\">\u751f\u6210\u5f0f\u4eba\u5de5\u667a\u6167<\/a>\u5982\u4f55\u5e6b\u52a9\u6700\u8907\u96dc\u7684\u5de5\u7a0b\u5de5\u4f5c\u4e4b\u4e00\uff1a\u8a2d\u8a08\u534a\u5c0e\u9ad4\u3002<\/p>\n\n\n\n<p>\u9019\u9805\u5de5\u4f5c\u5c55\u793a\u4e86\u5728\u9ad8\u5ea6\u5c08\u696d\u5316\u9818\u57df\u7684\u516c\u53f8\u5982\u4f55\u5229\u7528\u5167\u90e8\u8cc7\u6599\u8a13\u7df4\u5927\u578b\u8a9e\u8a00\u6a21\u578b (<a href=\"https:\/\/www.nvidia.com\/en-us\/glossary\/data-science\/large-language-models\/\" target=\"_blank\" rel=\"noreferrer noopener\">LLMs<\/a>)\uff0c\u4ee5\u5efa\u7acb\u63d0\u9ad8\u751f\u7522\u529b\u7684\u52a9\u624b\u3002<\/p>\n\n\n\n<p>\u534a\u5c0e\u9ad4\u8a2d\u8a08\u662f\u4e00\u9805\u6975\u5177\u6311\u6230\u6027\u7684\u5de5\u4f5c\u3002\u5728\u986f\u5fae\u93e1\u4e0b\uff0c\u50cf <a href=\"https:\/\/www.nvidia.com\/en-us\/data-center\/h100\/\" target=\"_blank\" rel=\"noreferrer noopener\">NVIDIA H100 Tensor Core GPU<\/a>\uff08\u4e0a\u5716\uff09\u9019\u6a23\u6700\u5148\u9032\u7684\u6676\u7247\u770b\u8d77\u4f86\u5c31\u50cf\u4e00\u5ea7\u7cbe\u5fc3\u898f\u5283\u7684\u5927\u90fd\u5e02\uff0c\u7531\u6578\u767e\u5104\u500b\u96fb\u6676\u9ad4\u69cb\u5efa\u800c\u6210\uff0c\u9023\u63a5\u5728\u6bd4\u4eba\u985e\u982d\u9aee\u7d30 10,000 \u500d\u7684\u8857\u9053\u4e0a\u3002<\/p>\n\n\n\n<p>\u591a\u500b\u5de5\u7a0b\u5718\u968a\u82b1\u8cbb\u9577\u9054\u5169\u5e74\u7684\u6642\u9593\u9032\u884c\u5354\u4f5c\uff0c\u4ee5\u5efa\u9020\u5176\u4e2d\u4e00\u500b\u9019\u6a23\u7684\u6578\u4f4d\u5316\u5de8\u578b\u57ce\u5e02\u3002<\/p>\n\n\n\n<p>\u6709\u4e9b\u5c0f\u7d44\u5b9a\u7fa9\u6676\u7247\u7684\u6574\u9ad4\u67b6\u69cb\uff0c\u6709\u4e9b\u5c0f\u7d44\u88fd\u4f5c\u4e26\u653e\u7f6e\u5404\u7a2e\u8d85\u5c0f\u578b\u96fb\u8def\uff0c\u6709\u4e9b\u5c0f\u7d44\u6e2c\u8a66\u4ed6\u5011\u7684\u5de5\u4f5c\u3002 \u6bcf\u9805\u5de5\u4f5c\u90fd\u9700\u8981\u5c08\u9580\u7684\u65b9\u6cd5\u3001\u8edf\u9ad4\u7a0b\u5f0f\u548c\u96fb\u8166\u8a9e\u8a00\u3002<\/p>\n\n\n\n<p><\/p>\n\n\n\n<p><strong>\u5c0d\u65bc\u5927\u578b\u8a9e\u8a00\u6a21\u578b\u7684\u5ee3\u6cdb\u9858\u666f<\/strong><\/p>\n\n\n\n<p>NVIDIA \u7814\u7a76\u7e3d\u76e3\u66a8\u8a72\u8ad6\u6587\u4e3b\u8981\u4f5c\u8005 Mark Ren \u8868\u793a\uff1a\u300c\u6211\u76f8\u4fe1\u96a8\u8457\u6642\u9593\u7684\u63a8\u79fb\uff0c\u5927\u578b\u8a9e\u8a00\u6a21\u578b\u5c07\u5168\u9762\u5e6b\u52a9\u6240\u6709\u6d41\u7a0b\u3002\u300d<\/p>\n\n\n\n<p>NVIDIA \u9996\u5e2d\u79d1\u5b78\u5bb6 Bill Dally \u4eca\u5929\u5728\u570b\u969b\u96fb\u8166\u8f14\u52a9\u8a2d\u8a08\u6703\u8b70\uff08International Conference on Computer-Aided Design\uff09\u4e0a\u7684\u4e3b\u984c\u6f14\u8b1b\u4e2d\u5ba3\u5e03\u4e86\u9019\u7bc7\u8ad6\u6587\u3002\u300c\u9019\u9805\u52aa\u529b\u4ee3\u8868\u8457\u5c07\u5927\u578b\u8a9e\u8a00\u6a21\u578b\u61c9\u7528\u65bc\u8907\u96dc\u7684\u534a\u5c0e\u9ad4\u8a2d\u8a08\u5de5\u4f5c\u4e2d\u9081\u51fa\u4e86\u91cd\u8981\u7684\u7b2c\u4e00\u6b65\u300d\uff0cDally \u5728\u820a\u91d1\u5c71\u8209\u884c\u7684\u6d3b\u52d5\u4e2d\u8868\u793a\uff1a\u300c\u5b83\u5c55\u793a\u4e86\u5373\u4f7f\u662f\u9ad8\u5ea6\u5c08\u696d\u5316\u7684\u9818\u57df\uff0c\u4e5f\u53ef\u4ee5\u5229\u7528\u5176\u5167\u90e8\u8cc7\u6599\u4f86\u8a13\u7df4\u6709\u7528\u7684\u751f\u6210\u4eba\u5de5\u667a\u6167\u6a21\u578b\u3002\u300d<\/p>\n\n\n\n<p><strong>ChipNeMo<\/strong><strong>\u6d6e\u51fa\u6c34\u9762<\/strong><\/p>\n\n\n\n<p>\u9019\u7bc7\u8ad6\u6587\u8a73\u7d30\u4ecb\u7d39\u4e86 NVIDIA \u5de5\u7a0b\u5e2b\u5011\u5982\u4f55\u70ba\u5167\u90e8\u4f7f\u7528\u5275\u5efa\u4e86\u4e00\u500b\u540d\u70ba ChipNeMo \u7684\u5ba2\u88fd\u5316\u5927\u578b\u8a9e\u8a00\u6a21\u578b\uff0c\u8a72\u6a21\u578b\u662f\u4f7f\u7528\u516c\u53f8\u5167\u90e8\u8cc7\u6599\u9032\u884c\u8a13\u7df4\uff0c\u4ee5\u751f\u6210\u548c\u512a\u5316\u8edf\u9ad4\uff0c\u4e26\u5354\u52a9\u4eba\u985e\u8a2d\u8a08\u5e2b\u3002<\/p>\n\n\n\n<p>\u9577\u671f\u4ee5\u4f86\uff0c\u5de5\u7a0b\u5e2b\u5011\u5e0c\u671b\u5c07\u751f\u6210\u5f0f\u4eba\u5de5\u667a\u6167\u61c9\u7528\u65bc\u6676\u7247\u8a2d\u8a08\u7684\u6bcf\u500b\u968e\u6bb5\uff0c\u9019\u6709\u53ef\u80fd\u986f\u8457\u63d0\u9ad8\u7e3d\u9ad4\u751f\u7522\u529b\uff0c\u5728 EDA \u9818\u57df\u5df2\u7d93\u6709\u8d85\u904e 20 \u5e74\u7d93\u9a57\u7684 Ren \u8868\u793a\u3002<\/p>\n\n\n\n<p>\u5728\u5c0d NVIDIA \u5de5\u7a0b\u5e2b\u9032\u884c\u53ef\u80fd\u7684\u7528\u4f8b\u8abf\u67e5\u5f8c\uff0c\u7814\u7a76\u5718\u968a\u9078\u64c7\u4e86\u4e09\u500b\u9805\u76ee\uff1a\u4e00\u500b\u804a\u5929\u6a5f\u5668\u4eba\u3001\u4e00\u500b\u7a0b\u5f0f\u78bc\u751f\u6210\u5668\u548c\u4e00\u500b\u5206\u6790\u5de5\u5177\u3002<\/p>\n\n\n\n<p><\/p>\n\n\n\n<p><strong>\u521d\u59cb\u7528\u4f8b<\/strong><strong><\/strong><\/p>\n\n\n\n<p>\u5f8c\u8005\uff08\u5206\u6790\u5de5\u5177\uff09\uff0c\u7528\u65bc\u81ea\u52d5\u5316\u8017\u6642\u7684\u4efb\u52d9\uff0c\u5373\u7dad\u8b77\u5df2\u77e5\u932f\u8aa4\u7684\u6700\u65b0\u63cf\u8ff0\uff0c\u76ee\u524d\u70ba\u6b62\u53d7\u5230\u4e86\u6700\u591a\u7684\u597d\u8a55\u3002<\/p>\n\n\n\n<p>\u4e00\u500b\u56de\u7b54\u6709\u95dc GPU \u67b6\u69cb\u548c\u8a2d\u8a08\u554f\u984c\u7684\u539f\u578b\u804a\u5929\u6a5f\u5668\u4eba\uff0c\u53ef\u5e6b\u52a9\u8a31\u591a\u5de5\u7a0b\u5e2b\u5728\u65e9\u671f\u6e2c\u8a66\u4e2d\u5feb\u901f\u627e\u5230\u6280\u8853\u6587\u4ef6\u3002<\/p>\n\n\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"600\" height=\"338\" src=\"https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2023\/10\/20231031-nvidia02.gif\" alt=\"\" class=\"wp-image-48001\" title=\"\"><figcaption class=\"wp-element-caption\">\u25b2\u7a0b\u5f0f\u78bc\u7522\u751f\u5668\u5c07\u5e6b\u52a9\u8a2d\u8a08\u4eba\u54e1\u70ba\u6676\u7247\u8a2d\u8a08\u7de8\u5beb\u8edf\u9ad4<\/figcaption><\/figure><\/div>\n\n\n<p><\/p>\n\n\n\n<p>\u6b63\u5728\u958b\u767c\u7684\u7a0b\u5f0f\u78bc\u7522\u751f\u5668\uff08\u5982\u4e0a\u65b9\u6f14\u793a\uff09\u5df2\u7d93\u900f\u904e\u6676\u7247\u8a2d\u8a08\u4eba\u54e1\u4f7f\u7528\u7684\u5169\u7a2e\u5c08\u7528\u8a9e\u8a00\uff0c\u5275\u5efa\u5927\u7d04 10-20 \u884c\u8edf\u9ad4\u7684\u7247\u6bb5\u3002\u5b83\u5c07\u8207\u73fe\u6709\u5de5\u5177\u6574\u5408\uff0c\u56e0\u6b64\u5de5\u7a0b\u5e2b\u5011\u5c07\u6709\u4e00\u500b\u65b9\u4fbf\u5feb\u6377\u7684\u8f14\u52a9\u5de5\u5177\u7528\u65bc\u9032\u884c\u4e2d\u7684\u8a2d\u8a08\u3002<\/p>\n\n\n\n<p><\/p>\n\n\n\n<p><strong>\u4f7f\u7528<\/strong><strong>NVIDIA NeMo<\/strong><strong>\u81ea\u8a02\u4eba\u5de5\u667a\u6167\u6a21\u578b<\/strong><\/p>\n\n\n\n<p>\u8a72\u8ad6\u6587\u4e3b\u8981\u5c08\u6ce8\u65bc\u5718\u968a\u6536\u96c6\u8a2d\u8a08\u8cc7\u6599\u4e26\u4f7f\u7528\u5176\u5275\u5efa\u5c08\u9580\u7684\u751f\u6210\u5f0f\u4eba\u5de5\u667a\u6167\u6a21\u578b\u7684\u5de5\u4f5c\uff0c\u9019\u662f\u4e00\u500b\u53ef\u8907\u88fd\u5230\u4efb\u4f55\u7522\u696d\u7684\u6d41\u7a0b\u3002<\/p>\n\n\n\n<p>\u4f5c\u70ba\u8d77\u9ede\uff0c\u5718\u968a\u9078\u64c7\u4e00\u500b<a href=\"https:\/\/blogs.nvidia.com\/blog\/2023\/03\/13\/what-are-foundation-models\/\" target=\"_blank\" rel=\"noreferrer noopener\">\u57fa\u790e\u6a21\u578b<\/a>\u4e26\u7528 <a href=\"https:\/\/www.nvidia.com\/en-us\/ai-data-science\/generative-ai\/nemo-framework\/\" target=\"_blank\" rel=\"noreferrer noopener\">NVIDIA NeMo<\/a> \u4f86\u5ba2\u88fd\u5316\u9019\u500b\u6a21\u578b\uff0cNVIDIA NeMo \u662f\u4e00\u500b\u7528\u65bc\u5efa\u7f6e\u3001\u81ea\u8a02\u548c\u90e8\u7f72\u751f\u6210\u5f0f\u4eba\u5de5\u667a\u6167\u6a21\u578b\u7684\u6846\u67b6\uff0c\u5b83\u5305\u542b\u5728 <a href=\"https:\/\/www.nvidia.com\/en-us\/data-center\/products\/ai-enterprise\/\" target=\"_blank\" rel=\"noreferrer noopener\">NVIDIA AI Enterprise<\/a> \u8edf\u9ad4\u5e73\u53f0\u4e2d\u3002 \u6240\u9078\u7684 NeMo \u6a21\u578b\u64c1\u6709 430 \u5104\u500b\u53c3\u6578\uff0c\u9019\u662f\u8861\u91cf\u5176\u7406\u89e3\u6a21\u5f0f\u80fd\u529b\u7684\u6307\u6a19\u3002\u5b83\u4f7f\u7528\u8d85\u904e\u4e00\u5146\u500b\u6b0a\u6756\uff08token\uff09\u9032\u884c\u4e86\u8a13\u7df4\uff0c\u9019\u4e9b\u6b0a\u6756\u4ee3\u8868\u6587\u672c\u548c\u8edf\u9ad4\u4e2d\u7684\u55ae\u8a5e\u548c\u7b26\u865f\u3002<\/p>\n\n\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"624\" height=\"313\" src=\"https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2023\/10\/20231031-nvidia03.jpg\" alt=\"\" class=\"wp-image-48002\" title=\"\" srcset=\"https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2023\/10\/20231031-nvidia03.jpg 624w, https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2023\/10\/20231031-nvidia03-300x150.jpg 300w\" sizes=\"(max-width: 624px) 100vw, 624px\" \/><figcaption class=\"wp-element-caption\">\u25b2ChipNeMo\u63d0\u4f9b\u4e86\u4e00\u500b\u7bc4\u4f8b\uff0c\u5c55\u793a\u6df1\u5ea6\u6280\u8853\u5718\u968a\u5982\u4f55\u5229\u7528\u5167\u90e8\u8cc7\u6599\u4f86\u6539\u9032\u9810\u8a13\u7df4\u6a21\u578b<\/figcaption><\/figure><\/div>\n\n\n<p><\/p>\n\n\n\n<p>\u96a8\u5f8c\uff0c\u5718\u968a\u5c0d\u6a21\u578b\u9032\u884c\u4e86\u5169\u8f2a\u8a13\u7df4\uff0c\u7b2c\u4e00\u8f2a\u4f7f\u7528\u4e86\u7d04 240 \u5104\u500b\u6b0a\u6756\u7684\u5167\u90e8\u8a2d\u8a08\u6578\u8cc7\u6599\uff0c\u7b2c\u4e8c\u8f2a\u4f7f\u7528\u4e86\u7d04 13 \u842c\u6b21\u5c0d\u8a71\u548c\u8a2d\u8a08\u793a\u4f8b\u7684\u6df7\u5408\u8cc7\u6599\u3002<\/p>\n\n\n\n<p>\u9019\u9805\u5de5\u4f5c\u662f\u534a\u5c0e\u9ad4\u7522\u696d\u4e2d\u751f\u6210\u5f0f\u4eba\u5de5\u667a\u6167\u7684<a href=\"https:\/\/arxiv.org\/pdf\/2308.10204.pdf\" target=\"_blank\" rel=\"noreferrer noopener\">\u7814\u7a76<\/a>\u548c<a href=\"https:\/\/www.cadence.com\/en_US\/home\/explore\/large-language-models-llms.html\" target=\"_blank\" rel=\"noreferrer noopener\">\u6982\u5ff5\u8b49\u660e<\/a>\u7684\u6578\u500b\u793a\u4f8b\u4e4b\u4e00\uff0c\u9019\u4e9b\u793a\u4f8b\u6b63\u5728\u9010\u6f38\u5f9e\u5be6\u9a57\u5ba4\u4e2d\u5d84\u9732\u982d\u89d2\u3002<\/p>\n\n\n\n<p><\/p>\n\n\n\n<p><strong>\u5206\u4eab\u6240\u5b78\u7d93\u9a57<\/strong><strong><\/strong><\/p>\n\n\n\n<p>Ren \u5206\u4eab\u4ed6\u548c\u5718\u968a\u5b78\u5230\u5176\u4e2d\u4e00\u500b\u5f88\u91cd\u8981\u7684\u7d93\u9a57\uff0c\u5373\u662f\u5ba2\u88fd\u5316\u5927\u578b\u8a9e\u8a00\u6a21\u578b\u7684\u50f9\u503c\u3002<\/p>\n\n\n\n<p>\u5728\u6676\u7247\u8a2d\u8a08\u4efb\u52d9\u4e2d\uff0c\u5177\u6709\u5c11\u81f3 130 \u5104\u500b\u53c3\u6578\u7684\u5ba2\u88fd\u5316 ChipNeMo \u6a21\u578b\u7684\u6548\u80fd\uff0c\u751a\u81f3\u53ef\u8207\u5177\u6709\u8d85\u904e 700 \u5104\u500b\u53c3\u6578\u7684 LLaMA2 \u7b49\u66f4\u5927\u7684\u901a\u7528\u578b\u5927\u578b\u8a9e\u8a00\u6a21\u578b\u7684\u6548\u80fd\u76f8\u5ab2\u7f8e\u6216\u8d85\u904e\u3002\u5728\u67d0\u4e9b\u7528\u4f8b\u4e2d\uff0cChipNeMo \u6a21\u578b\u8981\u597d\u5f97\u591a\u3002<\/p>\n\n\n\n<p>\u5728\u9019\u500b\u904e\u7a0b\u4e2d\uff0c\u4f7f\u7528\u8005\u9700\u8981\u8b39\u614e\u5730\u9078\u64c7\u4ed6\u5011\u6536\u96c6\u7684\u8cc7\u6599\uff0c\u4ee5\u53ca\u5982\u4f55\u6e05\u6d17\u9019\u4e9b\u8cc7\u6599\u4ee5\u4f9b\u8a13\u7df4\u4f7f\u7528\uff0c\u4ed6\u88dc\u5145\u8aaa\u3002<\/p>\n\n\n\n<p>\u6700\u5f8c\uff0cRen \u5efa\u8b70\u4f7f\u7528\u8005\u8981\u6642\u523b\u95dc\u6ce8\u53ef\u4ee5\u52a0\u901f\u548c\u7c21\u5316\u5de5\u4f5c\u7684\u6700\u65b0\u5de5\u5177\u3002<\/p>\n\n\n\n<p>NVIDIA \u7814\u7a76\u90e8\u9580\u64c1\u6709\u5168\u7403\u6578\u767e\u540d\u79d1\u5b78\u5bb6\u548c\u5de5\u7a0b\u5e2b\uff0c\u81f4\u529b\u65bc\u4eba\u5de5\u667a\u6167\u3001\u96fb\u8166\u5716\u5f62\u3001\u96fb\u8166\u8996\u89ba\u3001\u81ea\u99d5\u8eca\u548c\u6a5f\u5668\u4eba\u7b49\u9818\u57df\u7684\u7814\u7a76\u3002\u534a\u5c0e\u9ad4\u9818\u57df\u7684\u5176\u4ed6\u6700\u65b0\u9805\u76ee\u5305\u62ec\u4f7f\u7528\u4eba\u5de5\u667a\u6167\u8a2d\u8a08<a href=\"https:\/\/developer.nvidia.com\/blog\/designing-arithmetic-circuits-with-deep-reinforcement-learning\/\" target=\"_blank\" rel=\"noreferrer noopener\">\u66f4\u5c0f\u3001\u66f4\u5feb\u7684\u96fb\u8def<\/a>\uff0c\u4ee5\u53ca<a href=\"https:\/\/developer.nvidia.com\/blog\/autodmp-optimizes-macro-placement-for-chip-design-with-ai-and-gpus\/\" target=\"_blank\" rel=\"noreferrer noopener\">\u512a\u5316\u5927<\/a>\u578b\u5340\u584a\u7684\u5e03\u5c40\u3002<\/p>\n\n\n\n<p>\u5e0c\u671b\u5efa\u7acb\u81ea\u5df1\u7684\u5ba2\u88fd\u5316\u5927\u578b\u8a9e\u8a00\u6a21\u578b\u7684\u4f01\u696d\u73fe\u5728\u5c31\u53ef\u4ee5\u958b\u59cb\u4f7f\u7528 GitHub \u548c NVIDIA NGC \u76ee\u9304\u4e2d\u63d0\u4f9b\u7684 <a href=\"https:\/\/www.nvidia.com\/en-us\/ai-data-science\/products\/nemo\/get-started\/?nvid=nv-int-unbr-268853\" target=\"_blank\" rel=\"noreferrer noopener\">NeMo \u6846\u67b6<\/a>\u3002<\/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\n\n\n<p><\/p>\n","protected":false},"excerpt":{"rendered":"<p>\u4eca\u5929\u767c\u5e03\u7684\u4e00\u7bc7\u7814\u7a76\u8ad6\u6587\u63cf\u8ff0\u4e86\u751f\u6210\u5f0f\u4eba<\/p>\n","protected":false},"author":3,"featured_media":48000,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"rank_math_lock_modified_date":false,"footnotes":""},"categories":[13],"tags":[7660,71,6855,5315,7659],"class_list":["post-47999","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-news","tag-chipnemo","tag-nvidia","tag-nvidia-nemo","tag-5315","tag-7659"],"_links":{"self":[{"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/posts\/47999"}],"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=47999"}],"version-history":[{"count":4,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/posts\/47999\/revisions"}],"predecessor-version":[{"id":48006,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/posts\/47999\/revisions\/48006"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/media\/48000"}],"wp:attachment":[{"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=47999"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=47999"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=47999"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}