{"id":25195,"date":"2024-12-16T16:43:58","date_gmt":"2024-12-16T08:43:58","guid":{"rendered":"https:\/\/www.1ai.net\/?p=25195"},"modified":"2024-12-16T16:43:58","modified_gmt":"2024-12-16T08:43:58","slug":"%e6%97%a0%e9%97%ae%e8%8a%af%e7%a9%b9%e5%bc%80%e6%ba%90%e5%85%a8%e7%90%83%e9%a6%96%e6%ac%be%e7%ab%af%e4%be%a7%e5%85%a8%e6%a8%a1%e6%80%81%e7%90%86%e8%a7%a3%e6%a8%a1%e5%9e%8b-megrez-3b-omni%ef%bc%8c","status":"publish","type":"post","link":"https:\/\/www.1ai.net\/en\/25195.html","title":{"rendered":"No Questions Asked Core Dome Open Sources World's First End-Side Omnimodal Understanding Model Megrez-3B-Omni, Supports Image, Audio, and Text Understanding"},"content":{"rendered":"<p>December 16th.<a href=\"https:\/\/www.1ai.net\/en\/tag\/%e6%97%a0%e9%97%ae%e8%8a%af%e7%a9%b9\" title=\"[Sees articles with [no question core dome] label]\" target=\"_blank\" >No Question Core Dome<\/a>Announced today,<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>Megrez-3B-Omni, a full-modal understanding miniaturization model in the no-questions-asked core dome end-side solution, and its language-only model version, Megrez-3B-Instruct.<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-25196\" title=\"8d9ed1b9j00sokw7n00rpd000u00136p\" src=\"https:\/\/www.1ai.net\/wp-content\/uploads\/2024\/12\/8d9ed1b9j00sokw7n00rpd000u00136p.jpg\" alt=\"8d9ed1b9j00sokw7n00rpd000u00136p\" width=\"1080\" height=\"1410\" \/><\/p>\n<p>Officially, Megrez-3B-Omni is a full-modal understanding model made for the end, with the ability to process image, audio, and text modal data simultaneously:<\/p>\n<ul>\n<li>exist<strong>graphic understanding<\/strong>On the other hand, Megrez-3B-Omni is currently one of the most accurate image understanding models on several mainstream test sets such as OpenCompass, MME, MMMU, and OCRBench.<\/li>\n<li>exist<strong>text comprehension<\/strong>On the other hand, Megrez-3B-Omni achieves the optimal accuracy of the end-to-end model on several authoritative test sets such as C-EVAL, MMLU \/ MMLU Pro, AlignBench, and so on.<\/li>\n<li>exist<strong>speech understanding<\/strong>In terms of this, Megrez-3B-Omni supports voice input in both Chinese and English, and is also capable of handling complex multi-round dialog scenarios, as well as supporting voice questioning of input images or text, enabling free switching between different modes.<\/li>\n<\/ul>\n<p>Officials claim that the unimodal version of Megrez-3B-Instruct achieves a significant improvement in inference speed compared to its predecessor and other end-side macrolanguage models.<strong>Maximum inference speed can be ahead of the same precision model 300%<\/strong>.<\/p>\n<p>The relevant links are as follows:<\/p>\n<ul class=\"medium-size list-paddingleft-2\">\n<li>\n<p data-vmark=\"f313\"><strong>Github<\/strong>:<a href=\"https:\/\/github.com\/infinigence\/Infini-Megrez\" target=\"_blank\" rel=\"noopener\"><span class=\"link-text-start-with-http\">https:\/\/github.com\/infinigence\/Infini-Megrez<\/span><\/a><\/p>\n<\/li>\n<li>\n<p data-vmark=\"1ed6\"><strong>HuggingFace<\/strong>:<a href=\"https:\/\/huggingface.co\/Infinigence\/Megrez-3B-Omni\" target=\"_blank\" rel=\"noopener\"><span class=\"link-text-start-with-http\">https:\/\/huggingface.co\/Infinigence\/Megrez-3B-Omni<\/span><\/a><\/p>\n<\/li>\n<li>\n<p data-vmark=\"5227\"><strong>Infini-AI Heterogeneous Cloud<\/strong>:<a href=\"https:\/\/cloud.infini-ai.com\/genstudio\/model\/mo-c73owqiotql7lozr\" target=\"_blank\" rel=\"noopener\"><span class=\"link-text-start-with-http\">https:\/\/cloud.infini-ai.com\/genstudio\/model\/mo-c73owqiotql7lozr<\/span><\/a><\/p>\n<\/li>\n<li>\n<p data-vmark=\"89a4\"><strong>Modelers<\/strong>:<a href=\"https:\/\/modelers.cn\/models\/INFINIGENCE-AI\/Megrez-3B-Omni\" target=\"_blank\" rel=\"noopener\"><span class=\"link-text-start-with-http\">https:\/\/modelers.cn\/models\/INFINIGENCE-AI\/Megrez-3B-Omni<\/span><\/a><\/p>\n<\/li>\n<li>\n<p data-vmark=\"4445\"><strong>ModelScope<\/strong>:<a href=\"https:\/\/www.modelscope.cn\/models\/InfiniAI\/Megrez-3B-Omni\" target=\"_blank\" rel=\"noopener\"><span class=\"link-text-start-with-http\">https:\/\/www.modelscope.cn\/models\/InfiniAI\/Megrez-3B-Omni<\/span><\/a><\/p>\n<\/li>\n<\/ul>","protected":false},"excerpt":{"rendered":"<p>December 16th, 2011 - No Questions Asked Core Dome announced today that it has open sourced Megrez-3B-Omni, the full-modal comprehension mini-model of No Questions Asked Core Dome's end-side solution, and its language-only model version, Megrez-3B-Instruct. Officially, Megrez-3B-Omni is a full-modal comprehension model for the end-side, which is capable of processing three modalities of data: image comprehension and audio comprehension. Megrez-3B-Omni is a full-modal comprehension model built for endpoints, with the ability to process image, audio, and text at the same time: in terms of image comprehension, Megrez-3B-Omni is one of the image comprehension models with the highest accuracy on many mainstream test sets, such as OpenCompass, MME, MMMU, and OCRBench, and so on. In terms of text understanding, Megrez-3B-Omni is one of the most accurate models on C-E<\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[146],"tags":[219,2005],"collection":[],"class_list":["post-25195","post","type-post","status-publish","format-standard","hentry","category-news","tag-219","tag-2005"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/25195","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=25195"}],"version-history":[{"count":0,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/25195\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/media?parent=25195"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/categories?post=25195"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/tags?post=25195"},{"taxonomy":"collection","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/collection?post=25195"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}