{"id":27597,"date":"2025-01-22T21:52:15","date_gmt":"2025-01-22T13:52:15","guid":{"rendered":"https:\/\/www.1ai.net\/?p=27597"},"modified":"2025-01-22T21:52:15","modified_gmt":"2025-01-22T13:52:15","slug":"%e5%ad%97%e8%8a%82%e8%b7%b3%e5%8a%a8%e8%b1%86%e5%8c%85%e5%a4%a7%e6%a8%a1%e5%9e%8b-1-5-pro-%e5%8f%91%e5%b8%83%ef%bc%8c%e7%bb%bc%e5%90%88%e5%be%97%e5%88%86%e4%bc%98%e4%ba%8e-gpt-4o%e3%80%81claude-3-5-so","status":"publish","type":"post","link":"https:\/\/www.1ai.net\/en\/27597.html","title":{"rendered":"ByteDance Beanbag Big Model 1.5 Pro Released, Comprehensive Score Outperforms GPT-4o, Claude 3.5 Sonnet"},"content":{"rendered":"

January 22nd.ByteDance<\/a>UnderBean bag large model<\/a> 1.5 Pro is officially released today. The new Doubao-1.5-pro model has enhanced its comprehensive capabilities, with better overall scores than GPT-4o, Claude 3.5 Sonnet, and other models on multiple benchmarks, including knowledge, code, reasoning, and Chinese.<\/p>\n

\"b38fd961j00sqht5v0048d000u000rrp\"<\/p>\n

Currently, Doubao-1.5-pro has been gray-launched in Doubao App, and developers can also call the API directly from the volcano engine.<\/p>\n

According to the official introduction of ByteDance, the Beanbag Big Model 1.5 Pro uses smaller activation parameters for pre-training and adopts a large-scale sparse MoE architecture.Performance of Dense model with equivalent 7-fold activation parameters<\/strong>The MoE architecture is a very efficient and efficient system, far exceeding the conventional efficiency of the industry's MoE architectures by about 3x leverage.<\/p>\n

The new version of Doubao visual understanding model Doubao-1.5-vision-pro and Doubao real-time voice model Doubao-1.5-realtime-voice-pro are released together.<\/p>\n

Compared with the previous version, Doubao-1.5-vision-pro provides comprehensive technical upgrades in multimodal data synthesis, dynamic resolution, multimodal alignment, and hybrid training, which further enhances the model's capabilities in visual reasoning, text document recognition, fine-grained information understanding, and instruction following, and makes the model's response mode more streamlined and friendly.<\/p>\n

Beanbag Big Model 1.5 Pro This time, we have launched theBeanbag\u30fbReal-time speech modeling<\/strong>It is also available in the Doubao App (Doubao App has been upgraded to version 7.2.0). As a model integrating speech understanding and generation, Doubao real-time speech big model realizes end-to-end speech dialogue, with low latency, and can be interrupted at any time during the dialogue. Volcano Engine will launch the corresponding API service through the Ark platform in the first half of the year.<\/p>\n

1AI has learned from ByteDance officials that the Beanbag Grand Model 1.5 Pro declares \"All models were trained without using any other model-generated data<\/strong>\". Doubao Big Model 1.5 full products, including Doubao-1.5-pro, Doubao-1.5-lite, Doubao-1.5-vision-pro, etc., will continue to be priced the same as the original models, with no increase in quantity.<\/p>","protected":false},"excerpt":{"rendered":"

January 22 news, byte jump's Doubao big model 1.5 Pro officially released today, the new Doubao-1.5-pro model enhanced the comprehensive ability, in the knowledge, code, reasoning, Chinese and other benchmarks, the comprehensive score is better than the GPT-4o, Claude 3.5 Sonnet and other models. At present, Doubao-1.5-pro has been grayed out on line in Doubao App, while developers can also directly call the API in the volcano engine. According to the official introduction of ByteDance, Doubao-1.5-pro uses smaller activation parameters for pre-training, adopts large-scale sparse MoE architecture, and is equivalent to the performance of the Dense model with 7 times the activation parameters, which is far better than that of the industry.<\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[146],"tags":[548,2610],"collection":[],"class_list":["post-27597","post","type-post","status-publish","format-standard","hentry","category-news","tag-548","tag-2610"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/27597","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=27597"}],"version-history":[{"count":0,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/27597\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/media?parent=27597"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/categories?post=27597"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/tags?post=27597"},{"taxonomy":"collection","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/collection?post=27597"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}