{"id":6891,"date":"2025-06-17T05:22:00","date_gmt":"2025-06-16T20:22:00","guid":{"rendered":"https:\/\/devneko.jp\/wordpress\/?p=6891"},"modified":"2025-06-07T21:27:49","modified_gmt":"2025-06-07T12:27:49","slug":"the-common-pile-v0-1-an-8tb-dataset-of-public-domain-and-openly-licensed-text","status":"publish","type":"post","link":"https:\/\/devneko.jp\/wordpress\/?p=6891","title":{"rendered":"The Common Pile v0.1: An 8TB Dataset of Public Domain and Openly Licensed Text\u00a0"},"content":{"rendered":"\n<ul class=\"wp-block-list\">\n<li><strong>The Common Pile v0.1: An 8TB Dataset of Public Domain and Openly Licensed Text\u00a0<\/strong>[81.0]<br>\u30aa\u30fc\u30d7\u30f3\u30e9\u30a4\u30bb\u30f3\u30b9\u306e8\u30c6\u30e9\u30d0\u30a4\u30c8\u306e\u30c6\u30ad\u30b9\u30c8\u30b3\u30ec\u30af\u30b7\u30e7\u30f3\u3067\u3042\u308bCommon Pile v0.1\u3092\u53ce\u96c6\u3001\u30ad\u30e5\u30ec\u30fc\u30c8\u3001\u30ea\u30ea\u30fc\u30b9\u3057\u3066\u3044\u307e\u3059\u3002 Common Pile\u306f\u3001\u7814\u7a76\u8ad6\u6587\u3001\u30b3\u30fc\u30c9\u3001\u66f8\u7c4d\u3001\u767e\u79d1\u4e8b\u5178\u3001\u6559\u80b2\u8cc7\u6599\u3001\u30aa\u30fc\u30c7\u30a3\u30aa\u66f8\u304d\u8d77\u3053\u3057\u306a\u3069\u3001\u3055\u307e\u3056\u307e\u306a\u5206\u91ce\u306b\u307e\u305f\u304c\u308b30\u306e\u30bd\u30fc\u30b9\u304b\u3089\u306e\u30b3\u30f3\u30c6\u30f3\u30c4\u3067\u69cb\u6210\u3055\u308c\u3066\u3044\u308b\u3002 \u6211\u3005\u306f,\u30b3\u30e2\u30f3\u30d1\u30a4\u30eb\u304b\u3089\u30c6\u30ad\u30b9\u30c8\u306720\u5104\u306e\u30d1\u30e9\u30e1\u30fc\u30bfLSM\u3092\u30c8\u30ec\u30fc\u30cb\u30f3\u30b0\u3059\u308b\u3053\u3068\u3067,\u6211\u3005\u306e\u52aa\u529b\u3092\u691c\u8a3c\u3059\u308b\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2506.05209v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2506.05209v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Thu, 05 Jun 2025 16:21:30 GMT)<\/li>\n\n\n\n<li>\u300cWe release Common Pile v0.1, an 8TB corpus that\u2014to our knowledge\u2014constitutes the largest dataset built exclusively from openly licensed text. \u300d\u3068\u3044\u3046\u30af\u30ea\u30fc\u30f3\u306a\u30c7\u30fc\u30bf\u30bb\u30c3\u30c8\u69cb\u7bc9\u3068\u7af6\u4e89\u529b\u306e\u3042\u308b\u30e2\u30c7\u30eb\u69cb\u7bc9\u306e\u691c\u8a3c\u3002\u300cOur results demonstrate that not only is the Common Pile the strongest dataset for pretraining under an open-license constraint, but also that it produces models comparable to those trained on an equivalent amount of unlicensed data. This positive result holds promise for future of open-license pretraining, especially if the research community invests in collecting larger quantities of openly licensed text data in the future.\u300d\u3068\u306e\u3053\u3068\u3002<\/li>\n\n\n\n<li>\u975e\u5e38\u306b\u610f\u7fa9\u306e\u3042\u308b\u53d6\u308a\u7d44\u307f\u3060\u3068\u601d\u3046\u3001<\/li>\n\n\n\n<li>\u30c7\u30fc\u30bf\u30bb\u30c3\u30c8\u306f<a href=\"https:\/\/huggingface.co\/collections\/common-pile\/common-pile-v01-raw-data-6826b454a5a6a445d0b51b37\">Common Pile v0.1 Raw Data &#8211; a common-pile Collection<\/a>\u3001\u30ea\u30dd\u30b8\u30c8\u30ea\u306f<a href=\"https:\/\/github.com\/r-three\/common-pile\/tree\/main\">GitHub &#8211; r-three\/common-pile: Code for collecting, processing, and preparing datasets for the Common Pile<\/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":[491],"class_list":["post-6891","post","type-post","status-publish","format-standard","hentry","category-arxiv","tag-491"],"_links":{"self":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/6891","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=6891"}],"version-history":[{"count":1,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/6891\/revisions"}],"predecessor-version":[{"id":6892,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/6891\/revisions\/6892"}],"wp:attachment":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=6891"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=6891"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=6891"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}