{"id":4816,"date":"2024-05-06T04:13:00","date_gmt":"2024-05-05T19:13:00","guid":{"rendered":"https:\/\/devneko.jp\/wordpress\/?p=4816"},"modified":"2024-05-06T04:13:00","modified_gmt":"2024-05-05T19:13:00","slug":"prometheus-2-an-open-source-language-model-specialized-in-evaluating-other-language-models","status":"publish","type":"post","link":"https:\/\/devneko.jp\/wordpress\/?p=4816","title":{"rendered":"Prometheus 2: An Open Source Language Model Specialized in Evaluating Other Language Models"},"content":{"rendered":"\n<ul class=\"wp-block-list\">\n<li><strong>Prometheus 2: An Open Source Language Model Specialized in Evaluating Other Language Models\u00a0<\/strong>[92.7]<br>\u30d7\u30ed\u30e1\u30c6\u30a6\u30b92\u306f\u3001\u4eba\u9593\u3068GPT-4\u306e\u5224\u65ad\u3092\u5bc6\u63a5\u306b\u53cd\u6620\u3059\u308b\u3088\u308a\u5f37\u529b\u306a\u8a55\u4fa1\u5668\u3067\u3042\u308b\u3002 \u30e6\u30fc\u30b6\u5b9a\u7fa9\u8a55\u4fa1\u57fa\u6e96\u3067\u30b0\u30eb\u30fc\u30d7\u5316\u3055\u308c\u305f\u3001\u76f4\u63a5\u8a55\u4fa1\u3068\u30da\u30a2\u306e\u30e9\u30f3\u30ad\u30f3\u30b0\u30d5\u30a9\u30fc\u30de\u30c3\u30c8\u306e\u4e21\u65b9\u3092\u51e6\u7406\u3067\u304d\u308b\u3002 4\u3064\u306e\u76f4\u63a5\u8a55\u4fa1\u30d9\u30f3\u30c1\u30de\u30fc\u30af\u30684\u3064\u306e\u30da\u30a2\u306e\u30e9\u30f3\u30ad\u30f3\u30b0\u30d9\u30f3\u30c1\u30de\u30fc\u30af\u3067\u3001Prometheus 2\u306f\u4eba\u9593\u3068\u72ec\u81ea\u306eLM\u5224\u4e8b\u3068\u306e\u76f8\u95a2\u3068\u5408\u610f\u3092\u6700\u3082\u9ad8\u304f\u8a55\u4fa1\u3057\u3066\u3044\u308b\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2405.01535v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2405.01535v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Thu, 02 May 2024 17:59:35 GMT)<\/li>\n\n\n\n<li>\u8a55\u4fa1\u306e\u305f\u3081\u306eLM\u306e\u63d0\u6848\u3002GPT-4\u3092\u4f7f\u3063\u3066\u30c7\u30fc\u30bf\u3092\u69cb\u7bc9\u3001\u300cWe choose Mistral-7B (Jiang et al , 2023a) and Mixtral8x7B (Jiang et al , 2024) as our base models, and merge the weights of evaluator LMs separately trained on the FEEDBACK COLLECTION and the PREFERENCE COLLECTION to obtain our resulting models, PROMETHEUS 2 (7B &amp; 8x7B).\u300d\u3068\u306e\u3053\u3068\u3002<\/li>\n\n\n\n<li>\u30ea\u30dd\u30b8\u30c8\u30ea\u306f<a href=\"https:\/\/github.com\/prometheus-eval\/prometheus-eval\">GitHub &#8211; prometheus-eval\/prometheus-eval: Evaluate your LLM&#8217;s response with Prometheus \ud83d\udcaf<\/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":[223,648],"class_list":["post-4816","post","type-post","status-publish","format-standard","hentry","category-arxiv","tag-llm","tag-648"],"_links":{"self":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/4816","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=4816"}],"version-history":[{"count":0,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/4816\/revisions"}],"wp:attachment":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=4816"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=4816"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=4816"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}