{"id":75590,"date":"2024-07-24T15:41:38","date_gmt":"2024-07-24T07:41:38","guid":{"rendered":"https:\/\/www.ioiotimes.com\/?p=75590"},"modified":"2024-07-24T15:41:39","modified_gmt":"2024-07-24T07:41:39","slug":"nvidia-ai-foundry%e7%82%ba%e5%85%a8%e7%90%83%e4%bc%81%e6%a5%ad%e6%89%93%e9%80%a0%e5%ae%a2%e8%a3%bd%e5%8c%96llama-3-1%e7%94%9f%e6%88%90%e5%bc%8fai%e6%a8%a1%e5%9e%8b","status":"publish","type":"post","link":"https:\/\/www.ioiotimes.com\/?p=75590","title":{"rendered":"NVIDIA AI Foundry\u70ba\u5168\u7403\u4f01\u696d\u6253\u9020\u5ba2\u88fd\u5316Llama 3.1\u751f\u6210\u5f0fAI\u6a21\u578b"},"content":{"rendered":"\n<ul class=\"wp-block-list\">\n<li>\u4f01\u696d\u548c\u570b\u5bb6\u73fe\u5728\u53ef\u4ee5\u900f\u904e NVIDIA AI Foundry \u4f7f\u7528\u81ea\u5df1\u7684\u8cc7\u6599\u4e26\u642d\u914d Llama 3.1 405B \u548c NVIDIA Nemotron \u6a21\u578b\u4f86\u5efa\u7acb\u300c\u8d85\u7d1a\u6a21\u578b\u300d<em><\/em><\/li>\n\n\n\n<li>NVIDIA AI Foundry \u63d0\u4f9b\u5168\u9762\u7684\u751f\u6210\u5f0f AI \u6a21\u578b\u670d\u52d9\uff0c\u6db5\u84cb\u96c6\u5408\u7ba1\u7406\u3001\u5408\u6210\u8cc7\u6599\u7522\u751f\u3001\u5fae\u8abf\u3001\u6aa2\u7d22\u3001\u9632\u8b77\u548c\u8a55\u4f30\uff0c\u4ee5\u90e8\u7f72\u5177\u5099\u5168\u65b0NVIDIA NeMo Retriever\u5fae\u670d\u52d9\u4ee5\u63d0\u4f9b\u7cbe\u6e96\u56de\u61c9\u7684\u5ba2\u88fd\u5316Llama 3.1 NVIDIA NIM \u5fae\u670d\u52d9<em><\/em><\/li>\n\n\n\n<li>\u57c3\u68ee\u54f2\u7387\u5148\u4f7f\u7528\u65b0\u670d\u52d9\u70ba\u5ba2\u6236\u5efa\u69cb\u5ba2\u88fd\u5316Llama 3.1\u6a21\u578b\uff1b\u6c99\u70cf\u5730\u963f\u62c9\u4f2f\u570b\u5bb6\u77f3\u6cb9\u516c\u53f8\u3001AT&amp;T\u3001Uber \u548c\u5176\u4ed6\u7522\u696d\u9818\u5c0e\u8005\u7387\u5148\u4f7f\u7528\u5168\u65b0 Llama NVIDIA NIM \u5fae\u670d\u52d9<em><\/em><\/li>\n<\/ul>\n\n\n\n<p><\/p>\n\n\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"1280\" height=\"680\" src=\"https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2024\/07\/20240724-nvidia02.png\" alt=\"\" class=\"wp-image-75599\" title=\"\" srcset=\"https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2024\/07\/20240724-nvidia02.png 1280w, https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2024\/07\/20240724-nvidia02-300x159.png 300w, https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2024\/07\/20240724-nvidia02-1024x544.png 1024w, https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2024\/07\/20240724-nvidia02-768x408.png 768w\" sizes=\"(max-width: 1280px) 100vw, 1280px\" \/><\/figure><\/div>\n\n\n<p><\/p>\n\n\n\n<p>NVIDIA \u4eca\u5929\u5ba3\u5e03\u63a8\u51fa\u65b0\u7684 <a href=\"https:\/\/www.nvidia.com\/en-us\/ai\/foundry\/?ncid=ref-pr-188783\" target=\"_blank\" rel=\"noreferrer noopener\">NVIDIA AI Foundry<\/a> \u670d\u52d9\u548c <a href=\"https:\/\/www.nvidia.com\/zh-tw\/ai\/\" target=\"_blank\" rel=\"noreferrer noopener\">NVIDIA NIM<\/a>\u2122 \u63a8\u8ad6\u5fae\u670d\u52d9\uff0c\u900f\u904e\u4eca\u5929\u540c\u6a23\u63a8\u51fa\u7684 <a href=\"https:\/\/ai.meta.com\/blog\/meta-llama-3-1\/\" target=\"_blank\" rel=\"noreferrer noopener\">Llama 3.1<\/a> \u958b\u653e\u6a21\u578b\u7cfb\u5217\uff0c\u70ba\u5168\u7403\u4f01\u696d\u589e\u5f37\u751f\u6210\u5f0f\u4eba\u5de5\u667a\u6167\uff08AI\uff09\u80fd\u529b\u3002<\/p>\n\n\n\n<p>\u501f\u52a9 NVIDIA AI Foundry\uff0c\u4f01\u696d\u548c\u570b\u5bb6\u73fe\u5728\u53ef\u4ee5\u4f7f\u7528 Llama 3.1 \u548c NVIDIA \u8edf\u9ad4\u3001\u904b\u7b97\u548c\u5c08\u696d\u77e5\u8b58\u70ba\u5176\u7279\u5b9a\u9818\u57df\u7684\u7522\u696d\u7528\u4f8b\u5275\u5efa\u5ba2\u88fd\u5316\u300c\u8d85\u7d1a\u6a21\u578b\u300d\u3002\u4f01\u696d\u53ef\u4ee5\u4f7f\u7528\u5c08\u6709\u8cc7\u6599\u4ee5\u53ca Llama 3.1 405B \u548c <a href=\"https:\/\/blogs.nvidia.com\/blog\/nemotron-4-synthetic-data-generation-llm-training\/\" target=\"_blank\" rel=\"noreferrer noopener\">NVIDIA Nemotron<\/a>\u2122 Reward \u6a21\u578b\u7522\u751f\u7684\u5408\u6210\u8cc7\u6599\u4f86\u8a13\u7df4\u9019\u4e9b\u8d85\u7d1a\u6a21\u578b\u3002<\/p>\n\n\n\n<p>NVIDIA AI Foundry \u7531 <a href=\"https:\/\/www.nvidia.com\/en-us\/data-center\/dgx-cloud\/\" target=\"_blank\" rel=\"noopener\">NVIDIA <\/a><a href=\"https:\/\/www.nvidia.com\/en-us\/data-center\/dgx-cloud\/\" target=\"_blank\" rel=\"noreferrer noopener\">DGX<\/a><a href=\"https:\/\/www.nvidia.com\/zh-tw\/data-center\/dgx-cloud\/\" target=\"_blank\" rel=\"noopener\">\u2122 Cloud<\/a> AI \u5e73\u53f0\u9a45\u52d5\uff0c\u8a72\u5e73\u53f0\u662f\u8207\u4e16\u754c\u9818\u5148\u7684\u516c\u6709\u96f2\u5171\u540c\u8a2d\u8a08\u7684\uff0c\u53ef\u70ba\u4f01\u696d\u63d0\u4f9b\u91cd\u8981\u7684\u904b\u7b97\u8cc7\u6e90\uff0c\u4e26\u53ef\u6839\u64da AI \u9700\u6c42\u7684\u8b8a\u5316\u8f15\u9b06\u64f4\u5c55\u3002<\/p>\n\n\n\n<p>\u9019\u4e9b\u65b0\u7522\u54c1\u63a8\u51fa\u6b63\u503c\u4f01\u696d\u4ee5\u53ca\u5236\u5b9a\u4e3b\u6b0a AI \u7b56\u7565\u7684\u570b\u5bb6\u90fd\u5e0c\u671b\u5229\u7528\u7279\u5b9a\u9818\u57df\u7684\u77e5\u8b58\u4f86\u5efa\u7acb\u5ba2\u88fd\u5316\u7684\u5927\u578b\u8a9e\u8a00\u6a21\u578b\uff0c\u4ee5\u7528\u65bc\u53cd\u6620\u5176\u7368\u7279\u7684\u696d\u52d9\u6216\u6587\u5316\u7684\u751f\u6210\u5f0f AI \u61c9\u7528\u3002<\/p>\n\n\n\n<p>NVIDIA \u5275\u8fa6\u4eba\u66a8\u57f7\u884c\u9577\u9ec3\u4ec1\u52f3\u8868\u793a\uff1a\u300cMeta \u516c\u958b\u63a8\u51fa\u7684 Llama 3.1 \u6a21\u578b\u4ee3\u8868\u4e86\u5168\u7403\u4f01\u696d\u63a1\u7528\u751f\u6210\u5f0f AI \u7684\u95dc\u9375\u6642\u523b\u3002Llama 3.1 \u70ba\u6bcf\u500b\u4f01\u696d\u548c\u7522\u696d\u6253\u958b\u4e86\u5efa\u69cb\u6700\u5148\u9032\u7684\u751f\u6210\u5f0f\u4eba\u5de5\u667a\u6167\u61c9\u7528\u7684\u9598\u9580\u3002 NVIDIA AI Foundry \u5df2\u5168\u9762\u6574\u5408 Llama 3.1\uff0c\u4e26\u6e96\u5099\u597d\u5354\u52a9\u4f01\u696d\u5efa\u7f6e\u548c\u90e8\u7f72\u5ba2\u88fd\u5316 Llama \u8d85\u7d1a\u6a21\u578b\u3002\u300d<\/p>\n\n\n\n<p>Meta \u5275\u8fa6\u4eba\u66a8\u57f7\u884c\u9577\u99ac\u514b\uff0e\u7956\u514b\u4f2f\uff08Mark Zuckerberg\uff09\u8868\u793a\uff1a\u300c\u5168\u65b0\u7684 Llama 3.1 \u7cfb\u5217\u6a21\u578b\u70ba\u958b\u6e90 AI \u7684\u8d85\u7d1a\u91cd\u8981\u4e00\u6b65\u3002\u4f01\u696d\u900f\u904e NVIDIA AI Foundry \u53ef\u4ee5\u8f15\u9b06\u5275\u9020\u548c\u5ba2\u88fd\u5927\u5bb6\u6700\u60f3\u8981\u7684\u5148\u9032\u7684 AI \u670d\u52d9\uff0c\u4e26\u900f\u904e NVIDIA NIM \u90e8\u7f72\u3002\u6211\u5f88\u9ad8\u8208\u80fd\u628a\u5b83\u4ea4\u4ed8\u5230\u5927\u5bb6\u7684\u624b\u4e2d\u3002\u300d<\/p>\n\n\n\n<p>\u70ba\u4e86\u589e\u5f37\u7528\u65bc\u751f\u7522 AI \u7684 Llama 3.1 \u6a21\u578b\u7684\u4f01\u696d\u90e8\u7f72\uff0c\u73fe\u5728\u53ef\u5f9e <a href=\"http:\/\/ai.nvidia.com\" target=\"_blank\" rel=\"noreferrer noopener\">ai.nvidia.com<\/a>\u4e0b\u8f09\u9069\u7528\u65bc Llama 3.1 \u6a21\u578b\u7684 <a href=\"https:\/\/www.nvidia.com\/zh-tw\/ai\/\" target=\"_blank\" rel=\"noreferrer noopener\">NVIDIA NIM<\/a> \u63a8\u8ad6\u5fae\u670d\u52d9\u3002 NIM \u5fae\u670d\u52d9\u662f\u5728\u751f\u7522\u74b0\u5883\u4e2d\u90e8\u7f72 Llama 3.1 \u6a21\u578b\u7684\u6700\u5feb\u65b9\u5f0f\uff0c\u8207\u4e0d\u4f7f\u7528 NIM \u904b\u884c\u63a8\u7406\u76f8\u6bd4\uff0c\u541e\u5410\u91cf\u53ef\u63d0\u9ad8 2.5 \u500d\u3002<\/p>\n\n\n\n<p>\u4f01\u696d\u53ef\u4ee5\u5c07 Llama 3.1 NIM \u5fae\u670d\u52d9\u8207\u5168\u65b0 <a href=\"https:\/\/blogs.nvidia.com\/blog\/nemo-retriever-microservices\" target=\"_blank\" rel=\"noreferrer noopener\">NVIDIA NeMo Retriever NIM \u5fae\u670d\u52d9<\/a>\u914d\u5c0d\uff0c\u70ba AI copilots\u3001\u52a9\u7406\u548c<a href=\"https:\/\/www.nvidia.com\/zh-tw\/use-cases\/digital-humans\/\" target=\"_blank\" rel=\"noreferrer noopener\">\u6578\u4f4d\u4eba\u5206\u8eab<\/a>\u5275\u5efa\u6700\u5148\u9032\u7684\u6aa2\u7d22\u7ba1\u9053\u3002<\/p>\n\n\n\n<p><\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>\u57c3\u68ee\u54f2\u7387\u5148\u900f\u904eAI Foundry\u70ba\u4f01\u696d\u5ba2\u88fd\u5316Llama\u8d85\u7d1a\u6a21\u578b<\/strong><\/h3>\n\n\n\n<p>\u5168\u7403\u5c08\u696d\u670d\u52d9\u516c\u53f8 <a href=\"https:\/\/newsroom.accenture.com\/news\/2024\/accenture-pioneers-custom-llama-llm-models-with-nvidia-ai-foundry\" target=\"_blank\" rel=\"noreferrer noopener\">Accenture<\/a> \u7387\u5148\u63a1\u7528 NVIDIA AI Foundry\uff0c\u5229\u7528 Accenture AI Refinery\u2122 \u6846\u67b6\u5efa\u7acb\u5ba2\u88fd\u5316Llama 3.1 \u6a21\u578b\uff0c\u65e2\u4f9b\u5176\u81ea\u8eab\u4f7f\u7528\uff0c\u4e5f\u63d0\u4f9b\u5c0b\u6c42\u80fd\u5920\u90e8\u7f72\u53cd\u6620\u5176\u6587\u5316\u3001\u8a9e\u8a00\u548c\u7522\u696d\u7684\u751f\u6210\u5f0f AI \u61c9\u7528\u7a0b\u5f0f\u7684\u5ba2\u6236\u4f7f\u7528\u3002<\/p>\n\n\n\n<p>\u57c3\u68ee\u54f2\u8463\u4e8b\u66a8\u57f7\u884c\u9577\u6731\u8389\u00b7\u53f2\u5a01\u7279\uff08Julie Sweet\uff09\u8868\u793a\uff1a\u300c\u4e16\u754c\u9818\u5148\u7684\u4f01\u696d\u770b\u5230\u4e86\u751f\u6210\u5f0f AI \u5982\u4f55\u6539\u8b8a\u6bcf\u500b\u7522\u696d\uff0c\u4e26\u6e34\u671b\u90e8\u7f72\u7531\u5ba2\u88fd\u5316\u6a21\u578b\u9a45\u52d5\u7684\u61c9\u7528\u7a0b\u5f0f\u3002\u57c3\u68ee\u54f2\u4e00\u76f4\u65bc\u5167\u90e8 AI \u61c9\u7528\u4e2d\u4f7f\u7528 NVIDIA NIM \u63a8\u8ad6\u5fae\u670d\u52d9\uff0c\u73fe\u5728\u900f\u904e NVIDIA AI Foundry\uff0c\u6211\u5011\u53ef\u4ee5\u5e6b\u52a9\u5ba2\u6236\u5feb\u901f\u5275\u5efa\u548c\u90e8\u7f72\u5ba2\u88fd\u5316 Llama 3.1 \u6a21\u578b\uff0c\u9a45\u52d5\u70ba\u4ed6\u5011\u7684\u696d\u52d9\u76ee\u6a19\u8a2d\u8a08\u7684\u8b8a\u9769\u6027 AI \u61c9\u7528\u3002\u300d<\/p>\n\n\n\n<p>NVIDIA AI Foundry \u63d0\u4f9b\u7aef\u5c0d\u7aef\u670d\u52d9\uff0c\u53ef\u5feb\u901f\u5efa\u7acb\u5ba2\u88fd\u5316\u8d85\u7d1a\u6a21\u578b\u3002\u5b83\u5c07 NVIDIA \u8edf\u9ad4\u3001\u57fa\u790e\u8a2d\u65bd\u548c\u5c08\u696d\u77e5\u8b58\u8207 NVIDIA AI \u751f\u614b\u7cfb\u7684\u958b\u653e\u793e\u7fa4\u6a21\u5f0f\u3001\u6280\u8853\u548c\u652f\u63f4\u76f8\u7d50\u5408\u3002<\/p>\n\n\n\n<p>\u900f\u904e NVIDIA AI Foundry\uff0c\u4f01\u696d\u53ef\u4ee5\u4f7f\u7528 Llama 3.1 \u6a21\u578b\u548c <a href=\"https:\/\/www.nvidia.com\/zh-tw\/ai-data-science\/products\/nemo\/\" target=\"_blank\" rel=\"noreferrer noopener\">NVIDIA NeMo<\/a> \u5e73\u53f0\u5efa\u7acb\u5ba2\u88fd\u5316\u6a21\u578b\u3002<a href=\"https:\/\/www.nvidia.com\/zh-tw\/ai-data-science\/products\/nemo\/\" target=\"_blank\" rel=\"noreferrer noopener\">NVIDIA NeMo<\/a> \u5e73\u53f0\u5305\u62ec\u6709\u5728 <a href=\"https:\/\/huggingface.co\/spaces\/allenai\/reward-bench\" target=\"_blank\" rel=\"noreferrer noopener\">Hugging Face Reward Bench<\/a> \u4e0a\u6392\u540d\u7b2c\u4e00\u7684 NVIDIA Nemotron-4 340B Reward \u6a21\u578b\u3002<\/p>\n\n\n\n<p>\u4e00\u65e6\u5ba2\u88fd\u5316\u6a21\u578b\u5efa\u7acb\u5b8c\u6210\uff0c\u4f01\u696d\u53ef\u4ee5\u5728\u5176\u504f\u597d\u7684\u96f2\u7aef\u5e73\u53f0\u548c\u4f86\u81ea\u5168\u7403\u4f3a\u670d\u5668\u88fd\u9020\u5546\u7684 <a href=\"https:\/\/www.nvidia.com\/zh-tw\/data-center\/products\/certified-systems\/\" target=\"_blank\" rel=\"noreferrer noopener\">NVIDIA\u8a8d\u8b49\u7cfb\u7d71<\/a>\u4e0a\uff0c\u4f7f\u7528\u5176\u504f\u597d\u7684 MLOps \u548c AIOps \u5e73\u53f0\uff0c\u5275\u5efa NVIDIA NIM \u63a8\u8ad6\u5fae\u670d\u52d9\u65bc\u751f\u7522\u74b0\u5883\u4e2d\u4f86\u904b\u884c\u9019\u4e9b\u6a21\u578b\u3002<\/p>\n\n\n\n<p>NVIDIA AI Enterprise \u5c08\u5bb6\u548c\u5168\u7403\u7cfb\u7d71\u6574\u5408\u5546\u5925\u4f34\u8207 AI Foundry \u5ba2\u6236\u5408\u4f5c\uff0c\u52a0\u901f\u5f9e\u958b\u767c\u5230\u90e8\u7f72\u7684\u6574\u500b\u6d41\u7a0b\u3002<\/p>\n\n\n\n<p><\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>NVIDIA Nemotron\u63a8\u52d5\u5148\u9032\u6a21\u578b\u5ba2\u88fd\u5316<\/strong><\/h3>\n\n\n\n<p>\u9700\u8981\u984d\u5916\u8a13\u7df4\u8cc7\u6599\u4f86\u5275\u5efa\u7279\u5b9a\u9818\u57df\u6a21\u578b\u7684\u4f01\u696d\uff0c\u53ef\u4ee5\u540c\u6642\u4f7f\u7528 Llama 3.1 405B \u548c Nemotron-4 340B \u751f\u6210<a href=\"https:\/\/blogs.nvidia.com\/blog\/what-is-synthetic-data\/\" target=\"_blank\" rel=\"noreferrer noopener\">\u5408\u6210\u8cc7\u6599<\/a>\uff0c\u5f9e\u800c\u5728\u5275\u5efa\u5ba2\u88fd\u5316 Llama \u8d85\u7d1a\u6a21\u578b\u6642\u63d0\u9ad8\u6a21\u578b\u6e96\u78ba\u6027\u3002<\/p>\n\n\n\n<p>\u64c1\u6709\u81ea\u6709\u8a13\u7df4\u8cc7\u6599\u7684\u5ba2\u6236\uff0c\u53ef\u4ee5\u4f7f\u7528 <a href=\"https:\/\/www.nvidia.com\/zh-tw\/ai-data-science\/generative-ai\/nemo-framework\/\" target=\"_blank\" rel=\"noreferrer noopener\">NVIDIA NeMo<\/a> \u5ba2\u88fd\u5316 Llama 3.1 \u6a21\u578b\uff0c\u4ee5\u9032\u884c\u9818\u57df\u81ea\u9069\u61c9\u9810\u8a13\u7df4\uff08DAPT\uff09\uff0c\u4ee5\u9032\u4e00\u6b65\u63d0\u9ad8\u6a21\u578b\u6e96\u78ba\u6027\u3002<\/p>\n\n\n\n<p>NVIDIA \u548c Meta \u4e5f\u5408\u4f5c\u63d0\u4f9b\u4e86 Llama 3.1 \u7684\u84b8\u993e\u65b9\u6848\uff08distillation recipe\uff09\uff0c\u958b\u767c\u4eba\u54e1\u53ef\u4ee5\u85c9\u6b64\u69cb\u5efa\u66f4\u5c0f\u7684\u5ba2\u88fd\u5316 Llama 3.1 \u6a21\u578b\u4ee5\u7528\u65bc\u751f\u6210\u5f0f AI \u61c9\u7528\u3002\u9019\u4f7f\u4f01\u696d\u80fd\u5920\u5728\u66f4\u5ee3\u6cdb\u7684\u52a0\u901f\u57fa\u790e\u8a2d\u65bd\u4e0a\u904b\u884c Llama \u9a45\u52d5\u7684 AI \u61c9\u7528\uff0c\u4f8b\u5982 AI \u5de5\u4f5c\u7ad9\u548c\u7b46\u8a18\u578b\u96fb\u8166\u3002<\/p>\n\n\n\n<p><\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>\u696d\u754c\u9818\u5c0e\u4f01\u696d\u85c9\u52a9NVIDIA\u548cLlama\u589e\u5f37AI\u80fd\u529b<\/strong><\/h3>\n\n\n\n<p>\u91ab\u7642\u4fdd\u5065\u3001\u91d1\u878d\u670d\u52d9\u3001\u96f6\u552e\u3001\u904b\u8f38\u548c\u96fb\u4fe1\u9818\u57df\u7684\u516c\u53f8\u5df2\u7d93\u958b\u59cb\u4f7f\u7528 NVIDIA NIM \u5fae\u670d\u52d9\u4f86\u652f\u63f4 Llama\u3002\u9996\u6279\u4f7f\u7528 Llama 3.1 \u65b0\u7248 NIM \u5fae\u670d\u52d9\u7684\u4f01\u696d\u5305\u62ec\u6c99\u70cf\u5730\u963f\u62c9\u4f2f\u570b\u5bb6\u77f3\u6cb9\u516c\u53f8\u3001AT&amp;T\u3001Uber \u548c\u5176\u4ed6\u7522\u696d\u9818\u5c0e\u8005\u3002<\/p>\n\n\n\n<p>Llama 3.1 \u591a\u8a9e\u8a00\u5927\u578b\u8a9e\u8a00\u6a21\u578b\u96c6\u7fa4\u7531\u8d85\u904e 16,000 \u500b <a href=\"https:\/\/www.nvidia.com\/zh-tw\/data-center\/h100\/\" target=\"_blank\" rel=\"noreferrer noopener\">NVIDIA H100<\/a> Tensor Core GPU \u8a13\u7df4\u800c\u6210\uff0c\u4e26\u91dd\u5c0d NVIDIA \u52a0\u901f\u904b\u7b97\u548c\u8edf\u9ad4\u9032\u884c\u4e86\u6700\u4f73\u5316\u2014\u7121\u8ad6\u662f\u5728\u8cc7\u6599\u4e2d\u5fc3\u3001\u96f2\u7aef\u9084\u662f\u914d\u5099 <a href=\"https:\/\/www.nvidia.com\/zh-tw\/geforce\/rtx\/\" target=\"_blank\" rel=\"noopener\">NVIDIA RTX<\/a><a href=\"https:\/\/www.nvidia.com\/en-us\/geforce\/rtx\/\" target=\"_blank\" rel=\"noreferrer noopener\">\u2122 GPU<\/a> \u7684\u672c\u5730\u5de5\u4f5c\u7ad9\u6216\u914d\u5099 GeForce RTX GPU \u7684\u96fb\u8166\u4e0a\u3002\u9019\u4e9b\u751f\u6210\u5f0f AI \u6a21\u578b\u5305\u62ec 8B\u300170B \u548c 405B \u53c3\u6578\u898f\u6a21\u7684\u6a21\u578b\u3002<\/p>\n\n\n\n<p><\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>\u5168\u65b0NeMo Retriever RAG\u5fae\u670d\u52d9\u63d0\u5347\u6e96\u78ba\u6027\u548c\u6548\u80fd<\/strong><\/h3>\n\n\n\n<p>\u4f7f\u7528\u65b0\u7684 NVIDIA NeMo Retriever NIM \u63a8\u8ad6\u5fae\u670d\u52d9\u9032\u884c\u6aa2\u7d22\u589e\u5f37\u751f\u6210\uff08<a href=\"https:\/\/blogs.nvidia.com\/blog\/what-is-retrieval-augmented-generation\/\" target=\"_blank\" rel=\"noreferrer noopener\">RAG<\/a>\uff09\uff0c\u7d44\u7e54\u53ef\u4ee5\u5728\u751f\u7522\u74b0\u5883\u4e2d\u90e8\u7f72\u5ba2\u88fd\u5316\u7684 Llama \u8d85\u7d1a\u6a21\u578b\u548c Llama NIM \u5fae\u670d\u52d9\u6642\uff0c\u63d0\u5347\u56de\u61c9\u6e96\u78ba\u6027\u3002<\/p>\n\n\n\n<p>\u7d50\u5408 Llama 3.1 405B \u7684 NVIDIA NIM \u63a8\u8ad6\u5fae\u670d\u52d9\uff0cNeMo Retriever NIM \u5fae\u670d\u52d9\u70ba RAG \u7ba1\u7dda\u63d0\u4f9b\u6700\u9ad8\u7684\u958b\u6e90\u548c\u5546\u696d\u6587\u672c\u554f\u7b54\u6aa2\u7d22\u6e96\u78ba\u6027\u3002<\/p>\n\n\n\n<p><\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>\u4f01\u696d\u751f\u614b\u7cfb\u7d71\u6e96\u5099\u652f\u63f4Llama 3.1\u548cNeMo Retriever NIM\u90e8\u7f72<\/strong><\/h3>\n\n\n\n<p>\u6578\u767e\u5bb6\u63d0\u4f9b\u4f01\u696d\u3001\u8cc7\u6599\u548c\u57fa\u790e\u8a2d\u65bd\u5e73\u53f0\u7684 NVIDIA NIM \u5925\u4f34\uff0c\u73fe\u5728\u53ef\u4ee5\u5c07\u9019\u4e9b\u65b0\u5fae\u670d\u52d9\u6574\u5408\u5230\u4ed6\u5011\u7684 AI \u89e3\u6c7a\u65b9\u6848\u4e2d\uff0c\u70ba\u8d85\u904e 500 \u842c\u540d\u958b\u767c\u4eba\u54e1\u548c 19,000 \u5bb6\u65b0\u5275\u516c\u53f8\u7d44\u6210\u7684 NVIDIA \u793e\u7fa4\u589e\u5f37\u751f\u6210\u5f0f AI\u3002<\/p>\n\n\n\n<p><a href=\"https:\/\/www.nvidia.com\/zh-tw\/data-center\/products\/ai-enterprise\/\" target=\"_blank\" rel=\"noreferrer noopener\">NVIDIA AI Enterprise<\/a> \u73fe\u5df2\u63d0\u4f9b\u5c0d Llama 3.1 NIM \u548c NeMo Retriever NIM \u5fae\u670d\u52d9\u7684\u751f\u7522\u652f\u63f4\u3002<a href=\"https:\/\/developer.nvidia.com\/join-nvidia-developer-program\/\" target=\"_blank\" rel=\"noreferrer noopener\">NVIDIA \u958b\u767c\u8005\u8a08\u5283<\/a>\u7684\u6210\u54e1\u5f88\u5feb\u5c07\u80fd\u514d\u8cbb\u7372\u5f97 NIM \u5fae\u670d\u52d9\uff0c\u4ee5\u4fbf\u5728\u5176\u504f\u597d\u7684\u57fa\u790e\u8a2d\u65bd\u4e0a\u9032\u884c\u7814\u7a76\u3001\u958b\u767c\u548c\u6e2c\u8a66\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\u66f4\u591a<a href=\"https:\/\/www.ioiotimes.com\/?cat=9740\" data-type=\"link\" data-id=\"https:\/\/www.ioiotimes.com\/?cat=9740\" target=\"_blank\" rel=\"noreferrer noopener\">\u300aCOMPUTEX 2024\u300b<\/a>\u5831\u5c0e!!<br>\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>NVIDIA \u4eca\u5929\u5ba3\u5e03\u63a8\u51fa\u65b0\u7684 NV<\/p>\n","protected":false},"author":3,"featured_media":75597,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"rank_math_lock_modified_date":false,"footnotes":""},"categories":[13],"tags":[580,10362,71,10363],"class_list":["post-75590","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-news","tag-focus","tag-llama-3-1","tag-nvidia","tag-nvidia-ai-foundry"],"_links":{"self":[{"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/posts\/75590"}],"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=75590"}],"version-history":[{"count":4,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/posts\/75590\/revisions"}],"predecessor-version":[{"id":75601,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/posts\/75590\/revisions\/75601"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/media\/75597"}],"wp:attachment":[{"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=75590"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=75590"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=75590"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}