{"id":5062,"date":"2024-07-04T04:33:00","date_gmt":"2024-07-03T19:33:00","guid":{"rendered":"https:\/\/devneko.jp\/wordpress\/?p=5062"},"modified":"2024-07-04T04:33:00","modified_gmt":"2024-07-03T19:33:00","slug":"uio-llms-unbiased-incremental-optimization-for-long-context-llms","status":"publish","type":"post","link":"https:\/\/devneko.jp\/wordpress\/?p=5062","title":{"rendered":"UIO-LLMs: Unbiased Incremental Optimization for Long-Context LLMs"},"content":{"rendered":"\n<ul class=\"wp-block-list\">\n<li><strong>UIO-LLMs: Unbiased Incremental Optimization for Long-Context LLMs\u00a0<\/strong>[111.1]<br>UIO-LLMs\u306f\u3001\u9577\u3044\u30b3\u30f3\u30c6\u30ad\u30b9\u30c8\u8a2d\u5b9a\u4e0b\u3067\u306e\u30e1\u30e2\u30ea\u62e1\u5f35\u30c8\u30e9\u30f3\u30b9\u30d5\u30a9\u30fc\u30de\u30fc\u306e\u6f38\u9032\u7684\u306a\u6700\u9069\u5316\u624b\u6cd5\u3067\u3042\u308b\u3002 \u672c\u7a3f\u3067\u306f,TBPTT\u30a2\u30eb\u30b4\u30ea\u30ba\u30e0\u3092\u7528\u3044\u3066\u5b66\u7fd2\u904e\u7a0b\u3092\u6539\u826f\u3059\u308b\u3002 UIO-LLM\u306f\u3001Llama2-7b-chat\u306e\u30b3\u30f3\u30c6\u30ad\u30b9\u30c8\u30a6\u30a3\u30f3\u30c9\u30a6\u30924K\u304b\u3089100K\u30c8\u30fc\u30af\u30f3\u306b\u30012%\u306e\u8ffd\u52a0\u30d1\u30e9\u30e1\u30fc\u30bf\u3067\u62e1\u5f35\u3059\u308b\u306a\u3069\u3001\u9577\u3044\u30b3\u30f3\u30c6\u30ad\u30b9\u30c8\u3092\u6271\u3046\u3053\u3068\u306b\u6210\u529f\u3057\u305f\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2406.18173v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2406.18173v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Wed, 26 Jun 2024 08:44:36 GMT)<\/li>\n\n\n\n<li>\u30d9\u30fc\u30b9\u3068\u306a\u308bLLM\u3067\u30b3\u30f3\u30c6\u30f3\u30c4\u3092\u5727\u7e2e\u3001\u5727\u7e2e\u3057\u305f\u30b3\u30f3\u30c6\u30f3\u30c4\u3092\u4f7f\u3063\u3066\u30c7\u30b3\u30fc\u30c9\u3059\u308b\u3053\u3068\u3067\u9577\u6587\u3092\u53d6\u308a\u6271\u3046\u30a2\u30d7\u30ed\u30fc\u30c1\u306e\u63d0\u6848\u3002\uff08\u524d\u534a\u3068\u5f8c\u534a\u306e\u3064\u306a\u304e\u306b\u901a\u5e38\u306e\u30c6\u30ad\u30b9\u30c8\u3092\u4f7f\u3046\u4f8b\u306f\u3088\u304f\u898b\u308b\u306e\u3067\u30e2\u30c7\u30eb\u5185\u3067\u5b8c\u7d50\u3055\u305b\u3089\u308c\u3066\u3082\u4e0d\u601d\u8b70\u306f\u306a\u3044\u304c\u3001\u3061\u3083\u3093\u3068\u52d5\u4f5c\u3059\u308b\u3053\u3068\u306b\u9a5a\u304d\u304c\u3042\u308b\uff09<\/li>\n\n\n\n<li>\u30ea\u30dd\u30b8\u30c8\u30ea\u306f<a href=\"https:\/\/github.com\/wenhaoli-xmu\/UIO-LLMs\">GitHub &#8211; wenhaoli-xmu\/UIO-LLMs: Official implementation of UIO-LLMs<\/a><\/li>\n<\/ul>\n","protected":false},"excerpt":{"rendered":"","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[2],"tags":[223,667],"class_list":["post-5062","post","type-post","status-publish","format-standard","hentry","category-arxiv","tag-llm","tag-667"],"_links":{"self":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/5062","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=5062"}],"version-history":[{"count":0,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/5062\/revisions"}],"wp:attachment":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=5062"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=5062"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=5062"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}