{"id":6509,"date":"2025-03-31T04:04:00","date_gmt":"2025-03-30T19:04:00","guid":{"rendered":"https:\/\/devneko.jp\/wordpress\/?p=6509"},"modified":"2025-03-31T04:04:00","modified_gmt":"2025-03-30T19:04:00","slug":"scaling-laws-of-synthetic-data-for-language-models","status":"publish","type":"post","link":"https:\/\/devneko.jp\/wordpress\/?p=6509","title":{"rendered":"Scaling Laws of Synthetic Data for Language Models"},"content":{"rendered":"\n<ul class=\"wp-block-list\">\n<li><strong>Scaling Laws of Synthetic Data for Language Models\u00a0<\/strong>[132.7]<br>\u30d7\u30ec\u30c8\u30ec\u30fc\u30cb\u30f3\u30b0\u30b3\u30fc\u30d1\u30b9\u3092\u591a\u7a2e\u591a\u69d8\u306a\u9ad8\u54c1\u8cea\u306a\u5408\u6210\u30c7\u30fc\u30bf\u30bb\u30c3\u30c8\u306b\u5909\u63db\u3059\u308b\u30b9\u30b1\u30fc\u30e9\u30d6\u30eb\u306a\u30d5\u30ec\u30fc\u30e0\u30ef\u30fc\u30af\u3067\u3042\u308bSynthLLM\u3092\u7d39\u4ecb\u3057\u305f\u3002 \u63d0\u6848\u624b\u6cd5\u306f,\u30b0\u30e9\u30d5\u30a2\u30eb\u30b4\u30ea\u30ba\u30e0\u3092\u7528\u3044\u3066\u8907\u6570\u306e\u6587\u66f8\u306b\u307e\u305f\u304c\u308b\u30cf\u30a4\u30ec\u30d9\u30eb\u306a\u6982\u5ff5\u3092\u81ea\u52d5\u7684\u306b\u62bd\u51fa\u3057,\u518d\u7d50\u5408\u3059\u308b\u3053\u3068\u3067\u5b9f\u73fe\u3057\u3066\u3044\u308b\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2503.19551v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2503.19551v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Tue, 25 Mar 2025 11:07:12 GMT)<\/li>\n\n\n\n<li>\u5408\u6210\u30c7\u30fc\u30bf\u306eScaling law\u306b\u95a2\u3059\u308b\u5831\u544a\u3002\u9ad8\u54c1\u8cea\u306a\u30c7\u30fc\u30bf\u751f\u6210\u30d5\u30ec\u30fc\u30e0\u30ef\u30fc\u30afSYnathLLM\u3092\u524d\u63d0\u306b\u300cKey findings from our extensive mathematical experiments on SYNTHLLM include: (1) SYNTHLLM generates synthetic data that reliably adheres to the rectified scaling law across various model sizes; (2) Performance improvements plateau near 300B tokens; and (3) Larger models approach optimal performance with fewer training tokens.\u300d\u3068\u5408\u6210\u30c7\u30fc\u30bf\u306e\u6709\u52b9\u6027\u3092\u793a\u5506\u3059\u308b\u7d50\u8ad6\u306b\u306a\u3063\u3066\u3044\u308b\u3002<\/li>\n\n\n\n<li>\u30d7\u30ed\u30b8\u30a7\u30af\u30c8\u30b5\u30a4\u30c8\u306f<a href=\"https:\/\/thegenerality.com\/agi\/\">Advancing AI for Humanity<\/a>\u3002<\/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":[349,390],"class_list":["post-6509","post","type-post","status-publish","format-standard","hentry","category-arxiv","tag-scaling-law","tag-synthetic-data"],"_links":{"self":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/6509","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=6509"}],"version-history":[{"count":0,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/6509\/revisions"}],"wp:attachment":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=6509"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=6509"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=6509"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}