{"id":644,"date":"2025-04-08T14:54:41","date_gmt":"2025-04-08T14:54:41","guid":{"rendered":"https:\/\/www.juhenext.com\/?post_type=model&#038;p=644"},"modified":"2025-04-08T14:59:26","modified_gmt":"2025-04-08T14:59:26","slug":"bge-m3","status":"publish","type":"model","link":"https:\/\/www.juhenext.com\/zh\/model\/bge-m3\/","title":{"rendered":"bge-m3"},"content":{"rendered":"<p>BGE-M3 stands out for its <strong>Multi-Functionality<\/strong> (simultaneous dense, sparse, and multi-vector retrieval), <strong>Multi-Linguality<\/strong> (100+ languages), and <strong>Multi-Granularity<\/strong> (up to 8,192-token documents). It enhances retrieval pipelines by enabling hybrid retrieval (e.g., combining dense embeddings with BM25-like sparse weights) and re-ranking for higher accuracy. The model integrates seamlessly with tools like Vespa and Milvus, and its unified fine-tuning supports diverse retrieval methods. Recent updates include improved MIRACL benchmark performance and multilingual long-document datasets (MLDR).<\/p>","protected":false},"excerpt":{"rendered":"<p>BGE-M3 \u662f\u4e00\u4e2a\u591a\u529f\u80fd\u7684\u591a\u8bed\u8a00\u5d4c\u5165\u6a21\u578b\uff0c\u652f\u6301\u5728 100 \u591a\u79cd\u8bed\u8a00\u4e2d\u8fdb\u884c\u5bc6\u96c6\u3001\u7a00\u758f\u548c\u591a\u5411\u91cf\u68c0\u7d22\uff0c\u5e76\u5904\u7406\u4ece\u53e5\u5b50\u5230\u957f\u6587\u6863\uff088,192 \u4e2a\u6807\u8bb0\uff09\u7684\u8f93\u5165\u3002<\/p>","protected":false},"featured_media":556,"template":"","meta":{"_acf_changed":false},"context-window":[52],"features":[23],"maximum-output":[],"model-type":[25],"promotion":[],"provider":[53],"recommend":[],"class_list":["post-644","model","type-model","status-publish","has-post-thumbnail","hentry","context-window-8k","features-text-input","model-type-embedding","provider-opensource"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.juhenext.com\/zh\/wp-json\/wp\/v2\/model\/644","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.juhenext.com\/zh\/wp-json\/wp\/v2\/model"}],"about":[{"href":"https:\/\/www.juhenext.com\/zh\/wp-json\/wp\/v2\/types\/model"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.juhenext.com\/zh\/wp-json\/wp\/v2\/media\/556"}],"wp:attachment":[{"href":"https:\/\/www.juhenext.com\/zh\/wp-json\/wp\/v2\/media?parent=644"}],"wp:term":[{"taxonomy":"context-window","embeddable":true,"href":"https:\/\/www.juhenext.com\/zh\/wp-json\/wp\/v2\/context-window?post=644"},{"taxonomy":"features","embeddable":true,"href":"https:\/\/www.juhenext.com\/zh\/wp-json\/wp\/v2\/features?post=644"},{"taxonomy":"maximum-output","embeddable":true,"href":"https:\/\/www.juhenext.com\/zh\/wp-json\/wp\/v2\/maximum-output?post=644"},{"taxonomy":"model-type","embeddable":true,"href":"https:\/\/www.juhenext.com\/zh\/wp-json\/wp\/v2\/model-type?post=644"},{"taxonomy":"promotion","embeddable":true,"href":"https:\/\/www.juhenext.com\/zh\/wp-json\/wp\/v2\/promotion?post=644"},{"taxonomy":"provider","embeddable":true,"href":"https:\/\/www.juhenext.com\/zh\/wp-json\/wp\/v2\/provider?post=644"},{"taxonomy":"recommend","embeddable":true,"href":"https:\/\/www.juhenext.com\/zh\/wp-json\/wp\/v2\/recommend?post=644"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}