{"id":5256,"date":"2024-08-05T05:14:00","date_gmt":"2024-08-04T20:14:00","guid":{"rendered":"https:\/\/devneko.jp\/wordpress\/?p=5256"},"modified":"2024-08-05T05:14:00","modified_gmt":"2024-08-04T20:14:00","slug":"apple-intelligence-foundation-language-models","status":"publish","type":"post","link":"https:\/\/devneko.jp\/wordpress\/?p=5256","title":{"rendered":"Apple Intelligence Foundation Language Models"},"content":{"rendered":"\n<ul class=\"wp-block-list\">\n<li><strong>Apple Intelligence Foundation Language Models\u00a0<\/strong>[109.6]<br>\u672c\u5831\u544a\u3067\u306f\u3001\u30e2\u30c7\u30eb\u30a2\u30fc\u30ad\u30c6\u30af\u30c1\u30e3\u3001\u30e2\u30c7\u30eb\u30c8\u30ec\u30fc\u30cb\u30f3\u30b0\u306b\u4f7f\u7528\u3055\u308c\u308b\u30c7\u30fc\u30bf\u3001\u30c8\u30ec\u30fc\u30cb\u30f3\u30b0\u30d7\u30ed\u30bb\u30b9\u3001\u8a55\u4fa1\u7d50\u679c\u306b\u3064\u3044\u3066\u8ff0\u3079\u308b\u3002 \u79c1\u305f\u3061\u306f\u3001Responsible AI\u3068\u3001\u30e2\u30c7\u30eb\u958b\u767a\u5168\u4f53\u3092\u901a\u3057\u3066\u539f\u5247\u304c\u3069\u306e\u3088\u3046\u306b\u9069\u7528\u3055\u308c\u3066\u3044\u308b\u304b\u306b\u91cd\u70b9\u3092\u7f6e\u3044\u3066\u3044\u307e\u3059\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2407.21075v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2407.21075v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Mon, 29 Jul 2024 18:38:49 GMT)<\/li>\n\n\n\n<li>Apple\u306b\u3088\u308b\u57fa\u76e4\u30e2\u30c7\u30eb\u306e\u7d39\u4ecb\u3002\u300cAFM-server: We train AFM-server from scratch for 6.3T tokens on 8192 TPUv4 chips, using a sequence length of 4096 and a batch-size of 4096 sequences.\u300d\u3068\u3044\u3063\u305f\u3088\u3046\u306b\u304b\u306a\u308a\u8a73\u7d30\u306a\u5185\u5bb9\u304c\u8a18\u8f09\u3055\u308c\u3066\u3044\u308b\u3002\u300cAFM-on-device: For the on-device model, we found that knowledge distillation [Hinton et al , 2015] and structural pruning are effective ways to improve model performance and training efficiency.\u300d\u3068\u30c7\u30d0\u30a4\u30b9\u5411\u3051\u306f<a href=\"https:\/\/devneko.jp\/wordpress\/?p=5244\">MINITRON \/ Compact Language Models via Pruning and Knowledge Distillation \u2013 arXiv\u6700\u65b0\u8ad6\u6587\u306e\u7d39\u4ecb (devneko.jp)<\/a>\u3068\u8fd1\u3044\u30a2\u30d7\u30ed\u30fc\u30c1\u306b\u898b\u3048\u308b\u3002<\/li>\n\n\n\n<li>\u30d7\u30ed\u30b8\u30a7\u30af\u30c8\u30b5\u30a4\u30c8\u306f<a href=\"https:\/\/machinelearning.apple.com\/research\/introducing-apple-foundation-models\">Introducing Apple\u2019s On-Device and Server Foundation Models &#8211; Apple Machine Learning Research<\/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],"class_list":["post-5256","post","type-post","status-publish","format-standard","hentry","category-arxiv","tag-llm"],"_links":{"self":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/5256","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=5256"}],"version-history":[{"count":0,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/5256\/revisions"}],"wp:attachment":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=5256"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=5256"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=5256"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}