{"id":16330,"date":"2024-07-24T08:51:02","date_gmt":"2024-07-24T00:51:02","guid":{"rendered":"https:\/\/www.1ai.net\/?p=16330"},"modified":"2024-07-24T08:51:02","modified_gmt":"2024-07-24T00:51:02","slug":"%e6%89%8e%e5%85%8b%e4%bc%af%e6%a0%bc%ef%bc%9allama-3-1%e6%9c%89%e6%9c%9b%e6%88%90%e4%b8%ba%e5%bc%80%e6%ba%90ai%e7%9a%84linux","status":"publish","type":"post","link":"https:\/\/www.1ai.net\/en\/16330.html","title":{"rendered":"Zuckerberg: Llama 3.1 is expected to become the Linux of open source AI"},"content":{"rendered":"

Meta CEO MarkZuckerberg<\/a>Recently announcedLlama<\/a>3.1 is officially released, and it is said that this may beOpen Source AI<\/a>An important milestone in the field.<\/p>\n

Meta launched three Llama3.1 models of different sizes, most notably large models with 40.5 billion parameters. Zuckerberg described this as the \u201cstate-of-the-art open-source model\u201d that could even be compared to the leading closed-source model in some areas\u3002<\/p>\n

\"get-782\"<\/div>\n

In addition to the 405 billion parameter model, Meta also launched more streamlined 70 billion and 8 billion parameter models through knowledge distillation technology. Zuckerberg said that these two models have high cost performance while maintaining excellent performance.<\/p>\n

It is worth noting that Meta has updated its community policy to allow developers to use Llama3.1 as a teacher model for distillation and fine-tuning, which will undoubtedly greatly increase the application scope and potential of the model.<\/p>\n

When talking about the future of Llama, Zuckerberg compared it to another open source giant in the computer field, Linux. He reviewed the development history of Linux, pointing out that Linux initially gained the market due to its low cost and high customizability, and then eventually surpassed the closed source Unix system through a continuously improving ecosystem and extensive partner support.<\/p>\n

\u201cI think the Llama3.1 release was a similar turning point,\u201d said Zuckerberg, \u201cLlama has the opportunity to become an open-source AI industry standard.\u201d He stressed that, while some performance indicators might not have completely exceeded closed-source models, Llama had shown clear advantages in terms of cost, customization and fine-tuning of models\u3002<\/p>\n

Zuckerberg said Meta is working to build an ecosystem of partners to develop more features around Llama, and he is very excited about the future development prospects of Llama.<\/p>\n

With the rapid development of AI technology, the release of Llama 3.1 has undoubtedly injected new vitality into the open source AI community. The industry generally believes that this may accelerate the democratization of AI technology and enable more developers and companies to participate in AI innovation.<\/p>\n

Interested readers can visit the Llama official website (https:\/\/llama.meta.com) for more details.<\/p>\n

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

The Chief Executive Officer of Meta, Mark Zuckerberg, recently announced the official release of Llama3.1 and stated that this could be an important milestone in the open-source AI area. Meta launched three Llama3.1 models of different sizes, most notably large models with 40.5 billion parameters. Zuckerberg described this as the \u201cstate-of-the-art open-source model\u201d that could even be compared to the leading closed-source model in some areas. In addition to the 40.5 billion parametric model, Meta has introduced a more streamlined 70 billion and 8 billion parameter model through knowledge distillation technology. Zuckerberg stated that the two models, while maintaining excellent performance, had high value for money. It's worth noting that Meta update<\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[148,146],"tags":[184,3680,3679,296],"collection":[],"class_list":["post-16330","post","type-post","status-publish","format-standard","hentry","category-headline","category-news","tag-llama","tag-llama-3-1","tag-ai","tag-296"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/16330","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/comments?post=16330"}],"version-history":[{"count":0,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/16330\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/media?parent=16330"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/categories?post=16330"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/tags?post=16330"},{"taxonomy":"collection","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/collection?post=16330"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}