{"id":21392,"date":"2024-10-14T12:06:29","date_gmt":"2024-10-14T04:06:29","guid":{"rendered":"https:\/\/www.1ai.net\/?p=21392"},"modified":"2024-10-14T12:06:29","modified_gmt":"2024-10-14T04:06:29","slug":"%e6%99%ba%e8%b0%b1%e5%bc%80%e6%ba%90%e6%96%87%e7%94%9f%e5%9b%be%e6%a8%a1%e5%9e%8b-cogview3-plus%ef%bc%8c%e7%9b%b8%e5%85%b3%e5%8a%9f%e8%83%bd%e4%b8%8a%e7%ba%bf%e6%99%ba%e8%b0%b1%e6%b8%85%e8%a8%80-app","status":"publish","type":"post","link":"https:\/\/www.1ai.net\/en\/21392.html","title":{"rendered":"Wisdom Spectrum open source CogView3-Plus, related functions on the Wisdom Spectrum Clear Words App"},"content":{"rendered":"<p>October 14, 2012 - The Smart Spectrum technology team today announced that<a href=\"https:\/\/www.1ai.net\/en\/tag\/%e5%bc%80%e6%ba%90\" title=\"[View articles tagged with [open source]]\" target=\"_blank\" >Open Source<\/a><a href=\"https:\/\/www.1ai.net\/en\/tag\/%e6%96%87%e7%94%9f%e5%9b%be%e6%a8%a1%e5%9e%8b\" title=\"[Sees articles with tags]\" target=\"_blank\" >Wenshengtu Model<\/a><strong>\u00a0<\/strong><strong>CogView3 and CogView3-Plus-3B<\/strong><strong>\u00a0<\/strong>, the capabilities of the model series are now live\"<a href=\"https:\/\/www.1ai.net\/en\/tag\/%e6%99%ba%e8%b0%b1%e6%b8%85%e8%a8%80\" title=\"Look at the article that contains the tag\" target=\"_blank\" >Zhipu Qingyan<\/a>\"App.<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-21393\" title=\"9ca39406j00slbvcy01byd000u000tkm\" src=\"https:\/\/www.1ai.net\/wp-content\/uploads\/2024\/10\/9ca39406j00slbvcy01byd000u000tkm.jpg\" alt=\"9ca39406j00slbvcy01byd000u000tkm\" width=\"1080\" height=\"1064\" \/><\/p>\n<p>CogView3 is described as a text2img model based on cascade diffusion, which consists of the following three stages:<\/p>\n<ul>\n<li>Phase 1: Generate 512 x 512 low-resolution images using standard proliferation processes\u3002<\/li>\n<li>Phase II: Implementation of 2-fold ultra-resolution generation from 512 x 512 to generate 1024 x 1024 images using the trunking proliferation process\u3002<\/li>\n<li>Phase 3: The generated results are again iterated based on relay diffusion to generate 2048\u00d72048 high resolution images.<\/li>\n<\/ul>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-21394\" title=\"0384df8cj00slbvcx00f6d000u000fom\" src=\"https:\/\/www.1ai.net\/wp-content\/uploads\/2024\/10\/0384df8cj00slbvcx00f6d000u000fom.jpg\" alt=\"0384df8cj00slbvcx00f6d000u000fom\" width=\"1080\" height=\"564\" \/><\/p>\n<p>Officially, CogView3 outperforms the current state-of-the-art open-source text-to-image diffusion model SDXL by 77.01 TP3T in manual evaluation, while requiring only about 1\/10 of the inference time of SDXL.<\/p>\n<p>The CogView3-Plus model introduced the latest DiT framework based on CogView3 (ECCV'24) to further enhance overall performance. It was described as using Zero-SNR to spread noise and introduced<strong>Joint Text-Image Attention Mechanism<\/strong>CogView-3Plus uses a VAE with a potential dimension of 16. It effectively reduces training and inference costs while maintaining the basic capabilities of the model compared to the commonly used MMDiT structure.CogView-3Plus uses a VAE with a potential dimension of 16.<\/p>\n<p>The attached address is below:<\/p>\n<p data-vmark=\"9545\"><span class=\"referenceTitle\">Open source repository address:<\/span><\/p>\n<ul class=\"custom_reference list-paddingleft-1\">\n<li class=\"list-undefined list-reference-paddingleft\">\n<p data-vmark=\"0c7c\"><span class=\"link-text-start-with-http\">https:\/\/github.com\/THUDM\/CogView3<\/span><\/p>\n<\/li>\n<\/ul>\n<p data-vmark=\"2084\"><span class=\"referenceTitle\">Plus open source model repository:<\/span><\/p>\n<ul class=\"custom_reference list-paddingleft-1\">\n<li class=\"list-undefined list-reference-paddingleft\">\n<p data-vmark=\"a732\"><span class=\"link-text-start-with-http\">https:\/\/huggingface.co\/THUDM\/CogView3-Plus-3B<\/span><\/p>\n<\/li>\n<li class=\"list-undefined list-reference-paddingleft\">\n<p data-vmark=\"f21e\"><span class=\"link-text-start-with-http\">https:\/\/modelscope.cn\/models\/ZhipuAI\/CogView3-Plus-3B<\/span><\/p>\n<\/li>\n<\/ul>","protected":false},"excerpt":{"rendered":"<p>On October 14th, the SWAT team announced today the open source graphic models CogView3 and CogView3-Plus-3B, the capabilities of which are now on-line \"Intellection\" App. CogView3 is described as a text2img model based on cascade proliferation, which consists of the following three phases: first phase: 512 x 512 low-resolution images using the standard proliferation process. Phase II: Implementation of 2-fold ultra-resolution generation from 512 x 512 to generate 1024 x 1024 images using the trunking proliferation process. Phase III: Production results will again be based on re-extension, generating 2<\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[146],"tags":[219,1245,822],"collection":[],"class_list":["post-21392","post","type-post","status-publish","format-standard","hentry","category-news","tag-219","tag-1245","tag-822"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/21392","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=21392"}],"version-history":[{"count":0,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/21392\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/media?parent=21392"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/categories?post=21392"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/tags?post=21392"},{"taxonomy":"collection","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/collection?post=21392"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}