{"id":2479,"date":"2022-10-24T04:22:00","date_gmt":"2022-10-23T19:22:00","guid":{"rendered":"https:\/\/devneko.jp\/wordpress\/?p=2479"},"modified":"2022-10-24T04:22:00","modified_gmt":"2022-10-23T19:22:00","slug":"transcending-scaling-laws-with-0-1-extra-compute","status":"publish","type":"post","link":"https:\/\/devneko.jp\/wordpress\/?p=2479","title":{"rendered":"Transcending Scaling Laws with 0.1% Extra Compute\u00a0"},"content":{"rendered":"\n<ul class=\"wp-block-list\"><li><strong>Transcending Scaling Laws with 0.1% Extra Compute\u00a0<\/strong>[128.1]<br>\u8a00\u8a9e\u30e2\u30c7\u30eb\u3092\u30b9\u30b1\u30fc\u30eb\u3059\u308b\u3053\u3068\u3067\u30d1\u30d5\u30a9\u30fc\u30de\u30f3\u30b9\u304c\u5411\u4e0a\u3059\u308b\u304c\u3001\u8a08\u7b97\u30b3\u30b9\u30c8\u306f\u5927\u5e45\u306b\u5411\u4e0a\u3059\u308b\u3002 \u672c\u7a3f\u3067\u306f,\u65e2\u5b58\u306e\u8a00\u8a9e\u30e2\u30c7\u30eb\u3068\u305d\u306e\u30b9\u30b1\u30fc\u30ea\u30f3\u30b0\u66f2\u7dda\u3092,\u6bd4\u8f03\u7684\u5c11\u91cf\u306e\u4f59\u5270\u8a08\u7b97\u3067\u5927\u5e45\u306b\u6539\u5584\u3059\u308bUL2R\u3092\u63d0\u6848\u3059\u308b\u3002 \u307b\u307c\u7121\u8996\u53ef\u80fd\u306a\u4f59\u5206\u306a\u8a08\u7b97\u30b3\u30b9\u30c8\u3068\u65b0\u3057\u3044\u30c7\u30fc\u30bf\u30bd\u30fc\u30b9\u304c\u306a\u3051\u308c\u3070\u3001\u30c0\u30a6\u30f3\u30b9\u30c8\u30ea\u30fc\u30e0\u30e1\u30c8\u30ea\u30af\u30b9\u4e0a\u306e\u5927\u898f\u6a21\u8a00\u8a9e\u30e2\u30c7\u30eb\u306e\u30b9\u30b1\u30fc\u30ea\u30f3\u30b0\u7279\u6027\u3092\u5927\u5e45\u306b\u6539\u5584\u3067\u304d\u308b\u3053\u3068\u3092\u793a\u3059\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2210.11399v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2210.11399v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Thu, 20 Oct 2022 16:46:41 GMT)<ul><li>0.1 %\u3068\u5c11\u306a\u3044\u8a08\u7b97\u91cf\u3092\u4f7f\u7528\u3059\u308bUL2R\u3067\u6027\u80fd\u304c\u5927\u5e45\u306b\u5411\u4e0a\u3059\u308b\u3068\u306e\u8ad6\u6587<ul><li>\u8ad6\u6587\u4e2d\u306e\u8868\u8a18\u3060\u3068UL2: Uni\ufb01ed language learner Restore ? R: Restore?<\/li><li><a href=\"https:\/\/ai.googleblog.com\/2022\/10\/ul2-20b-open-source-unified-language.html\">UL2 20B: An Open Source Unified Language Learner \u2013 Google AI Blog (googleblog.com)<\/a>\u3067denoiser\u3092\u6df7\u5408\u3059\u308b\u30a2\u30d7\u30ed\u30fc\u30c1<\/li><li><a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2205.05131v1\">Fugu-MT \u8ad6\u6587\u7ffb\u8a33(\u6982\u8981): Unifying Language Learning Paradigms (fugumt.com)<\/a><\/li><\/ul><\/li><\/ul><ul><li>\u51e6\u7406\u3092\u884c\u3063\u305fU-PaLM\u306f\u3001\u82f1\u8a9e\u306eNLP\u30bf\u30b9\u30af(\u4f8b\u3048\u3070\u3001\u5e38\u8b58\u63a8\u8ad6\u3001\u8cea\u554f\u5fdc\u7b54)\u3001\u9023\u9396\u601d\u8003(eg\u3001GSM8K)\u3001\u591a\u8a00\u8a9e\u30bf\u30b9\u30af(MGSM\u3001TydiQA)\u3001MMLU\u3001BIG-Bench\u30bf\u30b9\u30af\u306a\u3069\u3067few-shot PaLM\u3088\u308a\u3082\u512a\u308c\u3066\u3044\u308b\u3068\u306e\u3053\u3068\u3002<\/li><\/ul><\/li><\/ul>\n","protected":false},"excerpt":{"rendered":"<p>Transcending Scaling Laws with 0.1% Extra Compute\u00a0[128.1]\u8a00\u8a9e\u30e2\u30c7\u30eb\u3092\u30b9\u30b1\u30fc\u30eb\u3059\u308b\u3053\u3068\u3067\u30d1\u30d5\u30a9\u30fc\u30de\u30f3\u30b9\u304c\u5411\u4e0a\u3059\u308b\u304c\u3001\u8a08\u7b97\u30b3\u30b9\u30c8\u306f\u5927\u5e45\u306b\u5411\u4e0a\u3059\u308b\u3002 \u672c\u7a3f\u3067\u306f,\u65e2\u5b58\u306e\u8a00 &hellip; <a href=\"https:\/\/devneko.jp\/wordpress\/?p=2479\" class=\"more-link\"><span class=\"screen-reader-text\">&#8220;Transcending Scaling Laws with 0.1% Extra Compute\u00a0&#8221; \u306e<\/span>\u7d9a\u304d\u3092\u8aad\u3080<\/a><\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[2],"tags":[170,634],"class_list":["post-2479","post","type-post","status-publish","format-standard","hentry","category-arxiv","tag-gigantic-language-model","tag-634"],"_links":{"self":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/2479","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=2479"}],"version-history":[{"count":0,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/2479\/revisions"}],"wp:attachment":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=2479"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=2479"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=2479"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}