{"id":5880,"date":"2024-12-12T06:05:00","date_gmt":"2024-12-11T21:05:00","guid":{"rendered":"https:\/\/devneko.jp\/wordpress\/?p=5880"},"modified":"2024-12-12T06:05:00","modified_gmt":"2024-12-11T21:05:00","slug":"nemotron-cc-transforming-common-crawl-into-a-refined-long-horizon-pretraining-dataset","status":"publish","type":"post","link":"https:\/\/devneko.jp\/wordpress\/?p=5880","title":{"rendered":"Nemotron-CC: Transforming Common Crawl into a Refined Long-Horizon Pretraining Dataset\u00a0"},"content":{"rendered":"\n<ul class=\"wp-block-list\">\n<li><strong>Nemotron-CC: Transforming Common Crawl into a Refined Long-Horizon Pretraining Dataset\u00a0<\/strong>[33.2]<br>\u7cbe\u5ea6\u3068\u30c7\u30fc\u30bf\u91cf\u3068\u306e\u30c8\u30ec\u30fc\u30c9\u30aa\u30d5\u3092\u6539\u5584\u3059\u308b\u65b9\u6cd5\u3092\u793a\u3057\u307e\u3059\u3002 15T\u30c8\u30fc\u30af\u30f3\u306e\u305f\u3081\u306b\u30c8\u30ec\u30fc\u30cb\u30f3\u30b0\u3055\u308c\u305f8B\u30d1\u30e9\u30e1\u30fc\u30bf\u30e2\u30c7\u30eb\u3067\u3001\u3046\u30617.2T\u306f\u3001Llama 3.1 8B\u30e2\u30c7\u30eb\u3088\u308a\u3082\u512a\u308c\u3066\u3044\u308b\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2412.02595v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2412.02595v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Tue, 03 Dec 2024 17:28:50 GMT)<\/li>\n\n\n\n<li>RedStone\u540c\u69d8\u3001Common Crawl\u3092\u3046\u307e\u304fRefine\u3059\u308b\u624b\u6cd5\u306e\u5831\u544a\u3002\u3053\u3061\u3089\u306fNDIVIA\u306b\u3088\u308b\u3082\u306e\u3002\u300cWe propose a method for transforming English Common Crawl into a 6.3T token longhorizon pretraining dataset, consisting of 4.4T globally deduplicated original tokens and 1.9T synthetically generated tokens.\u300d\u3068\u5408\u6210\u30c7\u30fc\u30bf\u306b\u3064\u3044\u3066\u89e6\u308c\u3089\u308c\u3066\u3044\u308b\u306e\u3082\u8208\u5473\u6df1\u3044\u3002<\/li>\n\n\n\n<li>\u30d7\u30ed\u30b8\u30a7\u30af\u30c8\u30b5\u30a4\u30c8\u306f<a href=\"https:\/\/data.commoncrawl.org\/contrib\/Nemotron\/Nemotron-CC\/index.html\">Nemotron-CC<\/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":[390,491],"class_list":["post-5880","post","type-post","status-publish","format-standard","hentry","category-arxiv","tag-synthetic-data","tag-491"],"_links":{"self":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/5880","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=5880"}],"version-history":[{"count":0,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/5880\/revisions"}],"wp:attachment":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=5880"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=5880"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=5880"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}