{"id":9026,"date":"2024-04-26T09:34:01","date_gmt":"2024-04-26T01:34:01","guid":{"rendered":"https:\/\/www.1ai.net\/?p=9026"},"modified":"2024-04-26T09:34:01","modified_gmt":"2024-04-26T01:34:01","slug":"%e8%82%a1%e4%bb%b7%e9%a3%99%e5%8d%87%ef%bc%81%e5%95%86%e6%b1%a4%e5%a4%a7%e6%a8%a1%e5%9e%8b%e6%8c%91%e6%88%98gpt4","status":"publish","type":"post","link":"https:\/\/www.1ai.net\/en\/9026.html","title":{"rendered":"Stock price soars! SenseTime&#039;s big model challenges GPT4"},"content":{"rendered":"<p>April 24,<a href=\"https:\/\/www.1ai.net\/en\/tag\/%e5%95%86%e6%b1%a4\" title=\"[Sees articles with [business soup] labels]\" target=\"_blank\" >SenseTime<\/a>The group&#039;s share price rose by 31.15% before the Hong Kong Stock Exchange suspended trading. SenseTime Group said that its<a href=\"https:\/\/www.1ai.net\/en\/tag\/%e5%a4%a7%e6%a8%a1%e5%9e%8b\" title=\"[View articles tagged with [large models]]\" target=\"_blank\" >Large Model<\/a>The launch of SenseTime 5.0 has attracted great attention from the market, and the company will publish further related announcements. At the &quot;2024 SenseTime Technology Exchange Day&quot;, SenseTime released its<span class=\"spamTxt\">up to date<\/span>The big model of China, SenseNova 5.0, aims to catch up with GPT-4, which reflects the collective goal of China&#039;s big model industry.<\/p>\n<p class=\"article-content__img\"><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-9027\" src=\"https:\/\/www.1ai.net\/wp-content\/uploads\/2024\/04\/6384966490206885763329225.png\" alt=\"\" width=\"720\" height=\"281\" \/><\/p>\n<p>By following the scaling laws of large language models, SenseNova 4.0 surpassed GPT-3.5 and SenseNova 5.0 fully matched GPT-4Turbo in less than three months. The scaling laws emphasize the power law relationship between model performance and model size, data volume, and computing power. SenseNova uses scientific experiments and mathematical formulas to predict the performance of the next generation of large models, avoiding blind attempts.<\/p>\n<p>Xu Li, Chairman and CEO of SenseTime, said that small-scale experiments can predict and verify the effectiveness of model architecture and data recipes, and ensure that these conclusions are maintained on a larger scale. SenseNova 5.0 uses more than 10 trillion tokens of pre-trained data in Chinese and English, and improves the model&#039;s reasoning, mathematics, and programming capabilities through logical synthetic data.<\/p>\n<p>SenseTime also launched the 1.8B parameter SenseChat-Lite, which surpassed all open source 2B models of the same level in mainstream evaluations. In addition, SenseTime released an SDK for end-side business and an enterprise-level large model all-in-one machine to support the data privatization deployment needs of industries such as finance, code, medical care, and government affairs.<\/p>\n<p>For software development, SenseTime released the lightweight version of the Raccoon Code large model all-in-one machine, which outperformed GPT-4 in the HumanEval test and supports multiple programming languages and contexts. It aims to help enterprise developers write, understand and maintain code more efficiently.<\/p>\n<p>SenseTime emphasized the importance of partners at the Technology Exchange Day and demonstrated the results of cooperation with companies such as Huawei Ascend. SenseTime&#039;s AI infrastructure SenseCore and AIDC computing power base provide strong support for it. SenseTime&#039;s generative AI business achieved significant growth in 2023, accounting for 35% of total revenue.<\/p>","protected":false},"excerpt":{"rendered":"<p>On 24 April, Comfort ' s share price increased by 31,151 TP3T prior to the suspension of trading at the Hong Kong Exchange. COMSO indicated that its large model day release of 5.0 would be of great interest to the market and that it would publish further related announcements. On the \u201cTechnology Exchange Day 2024\u201d, Comfort released its latest large model, SenseNoova5.0, aimed at catching up with GPT-4, which reflects the collective objectives of China's large model industry. In less than three months, through the Scaling Law, which follows a large-language model, the commercial soup has fully matched the GPT-4Turbo standard from Sensenova4.0 to GPT-3.5 to SenseNoova5.0. The scale law emphasizes model performance and model size, data volume and calculation<\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[146],"tags":[2391,216],"collection":[],"class_list":["post-9026","post","type-post","status-publish","format-standard","hentry","category-news","tag-2391","tag-216"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/9026","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=9026"}],"version-history":[{"count":0,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/9026\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/media?parent=9026"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/categories?post=9026"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/tags?post=9026"},{"taxonomy":"collection","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/collection?post=9026"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}