{"id":5816,"date":"2024-12-02T06:51:00","date_gmt":"2024-12-01T21:51:00","guid":{"rendered":"https:\/\/devneko.jp\/wordpress\/?p=5816"},"modified":"2024-12-02T06:51:00","modified_gmt":"2024-12-01T21:51:00","slug":"model-context-protocol-mcp-qwq-olmo-2","status":"publish","type":"post","link":"https:\/\/devneko.jp\/wordpress\/?p=5816","title":{"rendered":"Model Context Protocol (MCP), QwQ, OLMo 2"},"content":{"rendered":"\n<p>\u5148\u9031\u3082\u69d8\u3005\u306a\u30cb\u30e5\u30fc\u30b9\u304c\u3042\u3063\u305f\u304c\u3001\u6ce8\u76ee\u306fAnthropic\u306eModel Context Protocol\u3067\u3042\u308b\u3002\u3000<a href=\"https:\/\/www.anthropic.com\/news\/model-context-protocol\">Introducing the Model Context Protocol \\ Anthropic<\/a>\u3001<a href=\"https:\/\/modelcontextprotocol.io\/introduction\">Introduction &#8211; Model Context Protocol<\/a><\/p>\n\n\n\n<p>\u30b6\u30c3\u30af\u30ea\u3068\u306fLLM\u3068\u5916\u90e8\u30c7\u30fc\u30bf\u3084\u30c4\u30fc\u30eb\u3092\u7d71\u5408\u3059\u308b\u305f\u3081\u306e\u30d7\u30ed\u30c8\u30b3\u30eb\u3067\u3042\u308b\u3002\u5916\u90e8\u30c4\u30fc\u30eb\u5229\u7528\u3084\u30e1\u30e2\u30ea\u306e\u62e1\u5f35\u5229\u7528\u306a\u3069\u3092\u524d\u63d0\u3068\u3057\u305fLLM\u3092\u69cb\u7bc9\u3059\u308b\u5834\u5408\u3001\u3053\u306e\u624b\u306e\u6a19\u6e96\u304c\u3042\u308b\u304b\u306a\u3044\u304b\u306f\u91cd\u8981\u3002MCP\u304c\u30c7\u30d5\u30a1\u30af\u30c8\u30b9\u30bf\u30f3\u30c0\u30fc\u30c9\u3068\u306a\u308c\u308b\u304b\u8208\u5473\u6d25\u3005\u3002<\/p>\n\n\n\n<p>\u516c\u958b\u30e2\u30c7\u30eb\u95a2\u9023\u3067\u306f\u6975\u3081\u3066\u6027\u80fd\u306e\u9ad8\u3044Qwen with Questions\uff08QwQ\uff09\u3001\u4ee5\u524d\u53d6\u308a\u4e0a\u3052\u305f<a href=\"https:\/\/devneko.jp\/wordpress\/?p=4369\">Dolma\u3068OLMo \u2013 arXiv\u6700\u65b0\u8ad6\u6587\u306e\u7d39\u4ecb<\/a>\u306ever 2\u3067\u3042\u308bOLMo 2\u306b\u8981\u6ce8\u76ee\u3067\u3042\u308b\u3002O1 Replication Jurney\u3084TULU3\u3082\u3060\u304c\u3001\u3069\u306e\u3088\u3046\u306a\u624b\u6cd5\u3001\u30a2\u30d7\u30ed\u30fc\u30c1\u3067\u6027\u80fd\u304c\u4e0a\u304c\u308b\u306e\u304b\u306a\u3069\u3092\u30aa\u30fc\u30d7\u30f3\u306b\u3057\u305f\u53d6\u308a\u7d44\u307f\u306e\u4fa1\u5024\u306f\u9ad8\u3044\u3002<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><a href=\"https:\/\/qwenlm.github.io\/blog\/qwq-32b-preview\/\">QwQ: Reflect Deeply on the Boundaries of the Unknown | Qwen<\/a>\n<ul class=\"wp-block-list\">\n<li>\u300cQwQ-32B-Preview is an experimental research model developed by the Qwen Team, focused on advancing AI reasoning capabilities.\u300d\u3068\u3044\u3046\u516c\u958b\u30e2\u30c7\u30eb\u3002Open AI o1\u3068\u6bd4\u8f03\u3057\u3066\u3082\u6027\u80fd\u304c\u9ad8\u3044\u3002o1\u306b\u523a\u6fc0\u3092\u53d7\u3051\u305f\u52d5\u304d\u306f\u69d8\u3005\u884c\u308f\u308c\u3066\u3044\u3066\u672c\u5f53\u306b\u7af6\u4e89\u304c\u6fc0\u3057\u3044\u3002<\/li>\n\n\n\n<li>\u30ea\u30dd\u30b8\u30c8\u30ea\u306f<a href=\"https:\/\/huggingface.co\/Qwen\/QwQ-32B-Preview\">Qwen\/QwQ-32B-Preview \u00b7 Hugging Face<\/a><\/li>\n\n\n\n<li>\u30c7\u30e2\u306f<a href=\"https:\/\/huggingface.co\/spaces\/Qwen\/QwQ-32B-preview\">QwQ-32B-Preview &#8211; a Hugging Face Space by Qwen<\/a><\/li>\n<\/ul>\n<\/li>\n\n\n\n<li><a href=\"https:\/\/allenai.org\/blog\/olmo2\">OLMo 2: The best fully open language model to date | Ai2<\/a>\n<ul class=\"wp-block-list\">\n<li>\u69cb\u7bc9\u65b9\u6cd5\u3001\u30c7\u30fc\u30bf\u3001\u30e2\u30c7\u30eb\u304c\u516c\u958b\u3055\u308c\u3066\u3044\u308b\u30e2\u30c7\u30eb\u3067\u3042\u308a\u3001\u6027\u80fd\u306f\u6700\u5148\u7aef\u306b\u8fd1\u3044\u3002<\/li>\n\n\n\n<li>\u30ea\u30dd\u30b8\u30c8\u30ea\u306f<a href=\"https:\/\/huggingface.co\/collections\/allenai\/olmo-2-674117b93ab84e98afc72edc\">OLMo 2 &#8211; a allenai Collection<\/a><\/li>\n\n\n\n<li>\u30c7\u30e2\u306f<a href=\"https:\/\/playground.allenai.org\/\">Ai2 Playground<\/a><\/li>\n<\/ul>\n<\/li>\n<\/ul>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>O1 Replication Journey &#8212; Part 2: Surpassing O1-preview through Simple Distillation, Big Progress or Bitter Lesson?&nbsp;<\/strong>[30.9]<br>\u672c\u7a3f\u3067\u306f,OpenAI\u306eO1\u30e2\u30c7\u30eb\u6a5f\u80fd\u3092\u8907\u88fd\u3059\u308b\u73fe\u5728\u306e\u30a2\u30d7\u30ed\u30fc\u30c1\u306b\u3064\u3044\u3066,\u6279\u5224\u7684\u306a\u8003\u5bdf\u3092\u884c\u3046\u3002 O1\u306eAPI\u304b\u3089\u306e\u5358\u7d14\u306a\u84b8\u7559\u3068\u6559\u5e2b\u4ed8\u304d\u5fae\u8abf\u6574\u3092\u7d44\u307f\u5408\u308f\u305b\u308b\u3053\u3068\u3067\u3001\u8907\u96d1\u306a\u6570\u5b66\u7684\u63a8\u8ad6\u30bf\u30b9\u30af\u306b\u304a\u3044\u3066\u512a\u308c\u305f\u6027\u80fd\u304c\u5f97\u3089\u308c\u308b\u3053\u3068\u3092\u793a\u3059\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2411.16489v1\">\u8ad6\u6587<\/a>&nbsp;&nbsp;<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2411.16489v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>&nbsp; &nbsp;(Mon, 25 Nov 2024 15:31:27 GMT)<\/li>\n\n\n\n<li>OpenAI o1\u306b\u95a2\u3059\u308b\u7814\u7a76\u3001<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2410.18982v1\">Fugu-MT \u8ad6\u6587\u7ffb\u8a33(\u6982\u8981): O1 Replication Journey: A Strategic Progress Report &#8212; Part 1<\/a>\u304b\u3089\u306ePart2\u3002\u300cWhile our previous work (Part 1 (Qin et al , 2024)) explored the fundamental technical path to O1 replication, this study reveals how simple distillation from O1\u2019s API, combined with supervised fine-tuning, can achieve superior performance on complex mathematical reasoning tasks.\u300d\u306f\u307e\u3041\u3044\u3044\u3068\u3057\u3066\u300cNotably, despite training only on mathematical problem-solving data, our models demonstrated strong generalization to open-ended QA tasks and became significantly less susceptible to sycophancy after fine-tuning.\u300d\u306f\u9a5a\u304d\u3002<\/li>\n\n\n\n<li>\u30ea\u30dd\u30b8\u30c8\u30ea\u306f<a href=\"https:\/\/github.com\/GAIR-NLP\/O1-Journey\">GitHub &#8211; GAIR-NLP\/O1-Journey: O1 Replication Journey: A Strategic Progress Report \u2013 Part I<\/a><\/li>\n<\/ul>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>T\u00dcLU 3: Pushing Frontiers in Open Language Model Post-Training\u00a0<\/strong>[94.1]<br>\u6211\u3005\u306f\u3001\u5b8c\u5168\u306b\u30aa\u30fc\u30d7\u30f3\u306a\u6700\u5148\u7aef\u306e\u8a13\u7df4\u5f8c\u30e2\u30c7\u30eb\u3067\u3042\u308bT&#8221;ULU 3\u3092\u7d39\u4ecb\u3059\u308b\u3002 T&#8221;ULU 3\u306fLlama 3.1\u30d9\u30fc\u30b9\u30e2\u30c7\u30eb\u3092\u30d9\u30fc\u30b9\u306b\u3057\u3066\u304a\u308a\u3001Llama 3.1\u3001Qwen 2.5\u3001Mistral\u3001\u3055\u3089\u306bGPT-4o-mini\u3001Claude 3.5-Haiku\u3068\u3044\u3063\u305f\u30af\u30ed\u30fc\u30ba\u30c9\u30e2\u30c7\u30eb\u306b\u3082\u52dd\u3063\u3066\u3044\u308b\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2411.15124v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2411.15124v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Fri, 22 Nov 2024 18:44:04 GMT)<\/li>\n\n\n\n<li>\u30ea\u30dd\u30b8\u30c8\u30ea\u306f<a href=\"https:\/\/github.com\/allenai\/open-instruct\">GitHub &#8211; allenai\/open-instruct<\/a><\/li>\n<\/ul>\n","protected":false},"excerpt":{"rendered":"<p>\u5148\u9031\u3082\u69d8\u3005\u306a\u30cb\u30e5\u30fc\u30b9\u304c\u3042\u3063\u305f\u304c\u3001\u6ce8\u76ee\u306fAnthropic\u306eModel Context Protocol\u3067\u3042\u308b\u3002\u3000Introducing the Model Context Protocol \\ Anthropic\u3001Int &hellip; <a href=\"https:\/\/devneko.jp\/wordpress\/?p=5816\" class=\"more-link\"><span class=\"screen-reader-text\">&#8220;Model Context Protocol (MCP), QwQ, OLMo 2&#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":[42,223,251,283,293],"class_list":["post-5816","post","type-post","status-publish","format-standard","hentry","category-arxiv","tag-autonomous-agent","tag-llm","tag-mllm","tag-o1","tag-oss"],"_links":{"self":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/5816","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=5816"}],"version-history":[{"count":0,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/5816\/revisions"}],"wp:attachment":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=5816"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=5816"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=5816"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}