{"id":4762,"date":"2024-04-22T05:34:00","date_gmt":"2024-04-21T20:34:00","guid":{"rendered":"https:\/\/devneko.jp\/wordpress\/?p=4762"},"modified":"2024-04-22T05:34:00","modified_gmt":"2024-04-21T20:34:00","slug":"megalodon-efficient-llm-pretraining-and-inference-with-unlimited-context-length","status":"publish","type":"post","link":"https:\/\/devneko.jp\/wordpress\/?p=4762","title":{"rendered":"Megalodon: Efficient LLM Pretraining and Inference with Unlimited Context Length"},"content":{"rendered":"\n<ul class=\"wp-block-list\">\n<li><strong>Megalodon: Efficient LLM Pretraining and Inference with Unlimited Context Length\u00a0<\/strong>[112.8]<br>\u6587\u8108\u9577\u7121\u5236\u9650\u306e\u52b9\u7387\u7684\u306a\u30b7\u30fc\u30b1\u30f3\u30b9\u30e2\u30c7\u30ea\u30f3\u30b0\u306e\u305f\u3081\u306e\u30cb\u30e5\u30fc\u30e9\u30eb\u30cd\u30c3\u30c8\u30ef\u30fc\u30af\u3067\u3042\u308bMegalodon\u3092\u7d39\u4ecb\u3059\u308b\u3002 Llama2\u3068\u6bd4\u8f03\u3057\u3066\u3001Megalodon\u306f70\u5104\u306e\u30d1\u30e9\u30e1\u30fc\u30bf\u30682\u5146\u306e\u30c8\u30ec\u30fc\u30cb\u30f3\u30b0\u30c8\u30fc\u30af\u30f3\u306e\u30b9\u30b1\u30fc\u30eb\u3067Transformer\u3088\u308a\u3082\u52b9\u7387\u304c\u826f\u3044\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2404.08801v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2404.08801v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Fri, 12 Apr 2024 20:28:14 GMT)<\/li>\n\n\n\n<li>Transformer\u3088\u308a\u52b9\u7387\u304c\u826f\u3044\u3068\u3059\u308b\u69cb\u9020\u306e\u63d0\u6848\u3002MEGA (exponential moving average with gated attention)\u3092\u7d99\u627f\u3002\u540c\u898f\u6a21\u306eLlama2\u3088\u308a\u6027\u80fd\u304c\u3088\u3055\u305d\u3046\u3067\u9a5a\u304d\u3002<\/li>\n\n\n\n<li>\u30ea\u30dd\u30b8\u30c8\u30ea\u306f<a href=\"https:\/\/github.com\/XuezheMax\/megalodon\">XuezheMax\/megalodon: Reference implementation of Megalodon 7B model (github.com)<\/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":[223,241,551],"class_list":["post-4762","post","type-post","status-publish","format-standard","hentry","category-arxiv","tag-llm","tag-megalodon","tag-551"],"_links":{"self":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/4762","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=4762"}],"version-history":[{"count":0,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/4762\/revisions"}],"wp:attachment":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=4762"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=4762"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=4762"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}