{"id":11711,"date":"2024-04-24T11:04:46","date_gmt":"2024-04-24T11:04:46","guid":{"rendered":"https:\/\/dailyai.com\/?p=11711"},"modified":"2024-04-24T11:07:18","modified_gmt":"2024-04-24T11:07:18","slug":"microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm","status":"publish","type":"post","link":"https:\/\/dailyai.com\/da\/2024\/04\/microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm\/","title":{"rendered":"Microsoft lancerer Phi-3 Mini, en lille, men kraftfuld LM"},"content":{"rendered":"<p><strong>Microsoft har lanceret Phi-3 Mini, en lille sprogmodel, som er en del af virksomhedens strategi om at udvikle lette, funktionsspecifikke AI-modeller.<\/strong><\/p>\n<p>Udviklingen af sprogmodeller har set stadig st\u00f8rre parametre, tr\u00e6ningsdatas\u00e6t og kontekstvinduer. Skalering af st\u00f8rrelsen p\u00e5 disse modeller gav mere kraftfulde funktioner, men det havde en pris.<\/p>\n<p>Den traditionelle tilgang til at tr\u00e6ne en LLM er at lade den bruge enorme m\u00e6ngder data, hvilket kr\u00e6ver enorme computerressourcer. Tr\u00e6ning af en LLM som GPT-4 ansl\u00e5s f.eks. at have taget omkring 3 m\u00e5neder og kostet over $21 millioner.<\/p>\n<p>GPT-4 er en god l\u00f8sning til opgaver, der kr\u00e6ver komplekse r\u00e6sonnementer, men overkill til enklere opgaver som oprettelse af indhold eller en salgs-chatbot. Det er som at bruge en schweizerkniv, n\u00e5r alt, hvad du har brug for, er en simpel brev\u00e5bner.<\/p>\n<p>Med kun 3,8B parametre er Phi-3 Mini lille. Alligevel siger Microsoft, at det er en ideel, let og billig l\u00f8sning til opgaver som at opsummere et dokument, uddrage indsigter fra rapporter og skrive produktbeskrivelser eller indl\u00e6g p\u00e5 sociale medier.<\/p>\n<p>MMLU-benchmark-tallene viser, at Phi-3 Mini og de endnu ikke udgivne st\u00f8rre Phi-modeller sl\u00e5r st\u00f8rre modeller som <span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\">Mistral<\/span><\/span><\/span><\/span> 7B og Gemma 7B.<\/p>\n<figure id=\"attachment_11713\" aria-describedby=\"caption-attachment-11713\" style=\"width: 2000px\" class=\"wp-caption aligncenter\"><img loading=\"lazy\" decoding=\"async\" class=\"wp-image-11713 size-full\" src=\"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/The-Phi-3-small-language-models-with-big-potential-1.jpg\" alt=\"\" width=\"2000\" height=\"1125\" srcset=\"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/The-Phi-3-small-language-models-with-big-potential-1.jpg 2000w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/The-Phi-3-small-language-models-with-big-potential-1-300x169.jpg 300w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/The-Phi-3-small-language-models-with-big-potential-1-1024x576.jpg 1024w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/The-Phi-3-small-language-models-with-big-potential-1-768x432.jpg 768w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/The-Phi-3-small-language-models-with-big-potential-1-1536x864.jpg 1536w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/The-Phi-3-small-language-models-with-big-potential-1-18x10.jpg 18w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/The-Phi-3-small-language-models-with-big-potential-1-60x34.jpg 60w\" sizes=\"auto, (max-width: 2000px) 100vw, 2000px\" \/><figcaption id=\"caption-attachment-11713\" class=\"wp-caption-text\">Phi-3-modellernes pr\u00e6station p\u00e5 Massive Multitask Language Understanding (MMLU)-benchmark sammenlignet med andre modeller af samme st\u00f8rrelse. Kilde: Microsoft: Microsoft<\/figcaption><\/figure>\n<p>Microsoft siger, at Phi-3-small (7B parametre) og Phi-3-medium (14B parametre) vil v\u00e6re tilg\u00e6ngelige i Azure AI Model Catalog \"inden l\u00e6nge\".<\/p>\n<p>St\u00f8rre modeller som GPT-4 er stadig guldstandarden, og vi kan nok forvente, at GPT-5 bliver endnu st\u00f8rre.<\/p>\n<p>SLM'er som Phi-3 Mini har nogle vigtige fordele, som st\u00f8rre modeller ikke har. SLM'er er billigere at finjustere, kr\u00e6ver mindre beregning og kan k\u00f8re p\u00e5 enheden, selv i situationer, hvor der ikke er internetadgang.<\/p>\n<p>Implementering af en SLM p\u00e5 kanten resulterer i mindre ventetid og maksimal beskyttelse af personlige oplysninger, fordi der ikke er behov for at sende data frem og tilbage til skyen.<\/p>\n<p>Her er Sebastien Bubeck, VP for GenAI-forskning hos Microsoft AI, med en demonstration af Phi-3 Mini. Den er superhurtig og imponerende af s\u00e5 lille en model at v\u00e6re.<\/p>\n<blockquote class=\"twitter-tweet\">\n<p dir=\"ltr\" lang=\"en\">phi-3 er her, og den er ... god :-).<\/p>\n<p>Jeg har lavet en kort demo for at give dig en fornemmelse af, hvad phi-3-mini (3.8B) kan g\u00f8re. Hold \u00f8je med udgivelsen af \u00e5bne v\u00e6gte og flere meddelelser i morgen tidlig!<\/p>\n<p>(Og selvf\u00f8lgelig ville dette ikke v\u00e6re komplet uden den s\u00e6dvanlige tabel med benchmarks!) <a href=\"https:\/\/t.co\/AWA7Km59rp\">pic.twitter.com\/AWA7Km59rp<\/a><\/p>\n<p>- Sebastien Bubeck (@SebastienBubeck) <a href=\"https:\/\/twitter.com\/SebastienBubeck\/status\/1782627991874678809?ref_src=twsrc%5Etfw\">23. april 2024<\/a><\/p><\/blockquote>\n<p><script async src=\"https:\/\/platform.twitter.com\/widgets.js\" charset=\"utf-8\"><\/script><\/p>\n<h2>Kuraterede syntetiske data<\/h2>\n<p>Phi-3 Mini er et resultat af at forkaste ideen om, at store m\u00e6ngder data er den eneste m\u00e5de at tr\u00e6ne en model p\u00e5.<\/p>\n<p>Sebastien Bubeck, Microsofts vicepr\u00e6sident for generativ AI-forskning, spurgte: \"I stedet for bare at tr\u00e6ne p\u00e5 r\u00e5 webdata, hvorfor leder du s\u00e5 ikke efter data af ekstremt h\u00f8j kvalitet?\"<\/p>\n<p>Microsoft Researchs maskinl\u00e6ringsekspert Ronen Eldan l\u00e6ste godnathistorier for sin datter, da han spekulerede p\u00e5, om en sprogmodel kunne l\u00e6re ved kun at bruge ord, som en 4-\u00e5rig kunne forst\u00e5.<\/p>\n<p>Det f\u00f8rte til et eksperiment, hvor de skabte et datas\u00e6t, der startede med 3.000 ord. Ved kun at bruge dette begr\u00e6nsede ordforr\u00e5d fik de en LLM til at skabe millioner af korte b\u00f8rnehistorier, som blev samlet i et datas\u00e6t kaldet TinyStories.<\/p>\n<p>Forskerne brugte derefter TinyStories til at tr\u00e6ne en ekstremt lille model med 10 mio. parametre, som efterf\u00f8lgende var i stand til at generere \"flydende fort\u00e6llinger med perfekt grammatik\".<\/p>\n<p>De fortsatte med at iterere og skalere denne tilgang til generering af syntetiske data for at skabe mere avancerede, men omhyggeligt kuraterede og filtrerede syntetiske datas\u00e6t, som til sidst blev brugt til at tr\u00e6ne Phi-3 Mini.<\/p>\n<p>Resultatet er en lille model, der vil v\u00e6re mere overkommelig at k\u00f8re, samtidig med at den giver en ydelse, der kan sammenlignes med GPT-3.5.<\/p>\n<p>Mindre, men mere kompetente modeller vil f\u00e5 virksomheder til at bev\u00e6ge sig v\u00e6k fra blot at v\u00e6lge store LLM'er som GPT-4 som standard. Vi kan ogs\u00e5 snart se l\u00f8sninger, hvor en LLM h\u00e5ndterer de tunge l\u00f8ft, men uddelegerer enklere opgaver til letv\u00e6gtsmodeller.<\/p>","protected":false},"excerpt":{"rendered":"<p>Microsoft har lanceret Phi-3 Mini, en lille sprogmodel, som er en del af virksomhedens strategi om at udvikle lette, funktionsspecifikke AI-modeller. Udviklingen af sprogmodeller har set stadig st\u00f8rre parametre, tr\u00e6ningsdatas\u00e6t og kontekstvinduer. Ved at skalere st\u00f8rrelsen p\u00e5 disse modeller fik man mere kraftfulde funktioner, men det kostede. Den traditionelle tilgang til at tr\u00e6ne en LLM er at lade den bruge enorme m\u00e6ngder data, hvilket kr\u00e6ver enorme computerressourcer. Tr\u00e6ning af en LLM som GPT-4 ansl\u00e5s f.eks. at have taget omkring 3 m\u00e5neder og kostet over $21m. GPT-4 er en fantastisk l\u00f8sning til opgaver, der<\/p>","protected":false},"author":6,"featured_media":11715,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[84],"tags":[118,121],"class_list":["post-11711","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-industry","tag-llms","tag-microsoft"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v27.4 - https:\/\/yoast.com\/product\/yoast-seo-wordpress\/ -->\n<title>Microsoft launches Phi-3 Mini, a tiny but powerful LM | DailyAI<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/dailyai.com\/da\/2024\/04\/microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm\/\" \/>\n<meta property=\"og:locale\" content=\"da_DK\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Microsoft launches Phi-3 Mini, a tiny but powerful LM | DailyAI\" \/>\n<meta property=\"og:description\" content=\"Microsoft launched Phi-3 Mini, a tiny language model that is part of the company\u2019s strategy to develop lightweight, function-specific AI models. The progression of language models has seen ever larger parameters, training datasets, and context windows. Scaling the size of these models delivered more powerful capabilities but at a cost. The traditional approach to training an LLM is to have it consume massive amounts of data which requires huge computing resources. Training an LLM like GPT-4, for example, is estimated to have taken around 3 months and to have cost over $21m. GPT-4 is a great solution for tasks that\" \/>\n<meta property=\"og:url\" content=\"https:\/\/dailyai.com\/da\/2024\/04\/microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm\/\" \/>\n<meta property=\"og:site_name\" content=\"DailyAI\" \/>\n<meta property=\"article:published_time\" content=\"2024-04-24T11:04:46+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2024-04-24T11:07:18+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/tiny-AI.webp\" \/>\n\t<meta property=\"og:image:width\" content=\"1792\" \/>\n\t<meta property=\"og:image:height\" content=\"1024\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/webp\" \/>\n<meta name=\"author\" content=\"Eugene van der Watt\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:creator\" content=\"@DailyAIOfficial\" \/>\n<meta name=\"twitter:site\" content=\"@DailyAIOfficial\" \/>\n<meta name=\"twitter:label1\" content=\"Skrevet af\" \/>\n\t<meta name=\"twitter:data1\" content=\"Eugene van der Watt\" \/>\n\t<meta name=\"twitter:label2\" content=\"Estimeret l\u00e6setid\" \/>\n\t<meta name=\"twitter:data2\" content=\"3 minutter\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"NewsArticle\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm\\\/#article\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm\\\/\"},\"author\":{\"name\":\"Eugene van der Watt\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/#\\\/schema\\\/person\\\/7ce525c6d0c79838b7cc7cde96993cfa\"},\"headline\":\"Microsoft launches Phi-3 Mini, a tiny but powerful LM\",\"datePublished\":\"2024-04-24T11:04:46+00:00\",\"dateModified\":\"2024-04-24T11:07:18+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm\\\/\"},\"wordCount\":652,\"publisher\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/#organization\"},\"image\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2024\\\/04\\\/tiny-AI.webp\",\"keywords\":[\"LLMS\",\"Microsoft\"],\"articleSection\":[\"Industry\"],\"inLanguage\":\"da-DK\"},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm\\\/\",\"url\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm\\\/\",\"name\":\"Microsoft launches Phi-3 Mini, a tiny but powerful LM | DailyAI\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm\\\/#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2024\\\/04\\\/tiny-AI.webp\",\"datePublished\":\"2024-04-24T11:04:46+00:00\",\"dateModified\":\"2024-04-24T11:07:18+00:00\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm\\\/#breadcrumb\"},\"inLanguage\":\"da-DK\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm\\\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"da-DK\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm\\\/#primaryimage\",\"url\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2024\\\/04\\\/tiny-AI.webp\",\"contentUrl\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2024\\\/04\\\/tiny-AI.webp\",\"width\":1792,\"height\":1024},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/dailyai.com\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Microsoft launches Phi-3 Mini, a tiny but powerful LM\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/#website\",\"url\":\"https:\\\/\\\/dailyai.com\\\/\",\"name\":\"DailyAI\",\"description\":\"Your Daily Dose of AI News\",\"publisher\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/dailyai.com\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"da-DK\"},{\"@type\":\"Organization\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/#organization\",\"name\":\"DailyAI\",\"url\":\"https:\\\/\\\/dailyai.com\\\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"da-DK\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/#\\\/schema\\\/logo\\\/image\\\/\",\"url\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2023\\\/06\\\/Daily-Ai_TL_colour.png\",\"contentUrl\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2023\\\/06\\\/Daily-Ai_TL_colour.png\",\"width\":4501,\"height\":934,\"caption\":\"DailyAI\"},\"image\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/#\\\/schema\\\/logo\\\/image\\\/\"},\"sameAs\":[\"https:\\\/\\\/x.com\\\/DailyAIOfficial\",\"https:\\\/\\\/www.linkedin.com\\\/company\\\/dailyaiofficial\\\/\",\"https:\\\/\\\/www.youtube.com\\\/@DailyAIOfficial\"]},{\"@type\":\"Person\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/#\\\/schema\\\/person\\\/7ce525c6d0c79838b7cc7cde96993cfa\",\"name\":\"Eugene van der Watt\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"da-DK\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2023\\\/07\\\/Eugine_Profile_Picture-96x96.png\",\"url\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2023\\\/07\\\/Eugine_Profile_Picture-96x96.png\",\"contentUrl\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2023\\\/07\\\/Eugine_Profile_Picture-96x96.png\",\"caption\":\"Eugene van der Watt\"},\"description\":\"Eugene comes from an electronic engineering background and loves all things tech. When he takes a break from consuming AI news you'll find him at the snooker table.\",\"sameAs\":[\"www.linkedin.com\\\/in\\\/eugene-van-der-watt-16828119\"],\"url\":\"https:\\\/\\\/dailyai.com\\\/da\\\/author\\\/eugene\\\/\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Microsoft lancerer Phi-3 Mini, en lille, men kraftfuld LM | DailyAI","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/dailyai.com\/da\/2024\/04\/microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm\/","og_locale":"da_DK","og_type":"article","og_title":"Microsoft launches Phi-3 Mini, a tiny but powerful LM | DailyAI","og_description":"Microsoft launched Phi-3 Mini, a tiny language model that is part of the company\u2019s strategy to develop lightweight, function-specific AI models. The progression of language models has seen ever larger parameters, training datasets, and context windows. Scaling the size of these models delivered more powerful capabilities but at a cost. The traditional approach to training an LLM is to have it consume massive amounts of data which requires huge computing resources. Training an LLM like GPT-4, for example, is estimated to have taken around 3 months and to have cost over $21m. GPT-4 is a great solution for tasks that","og_url":"https:\/\/dailyai.com\/da\/2024\/04\/microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm\/","og_site_name":"DailyAI","article_published_time":"2024-04-24T11:04:46+00:00","article_modified_time":"2024-04-24T11:07:18+00:00","og_image":[{"width":1792,"height":1024,"url":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/tiny-AI.webp","type":"image\/webp"}],"author":"Eugene van der Watt","twitter_card":"summary_large_image","twitter_creator":"@DailyAIOfficial","twitter_site":"@DailyAIOfficial","twitter_misc":{"Skrevet af":"Eugene van der Watt","Estimeret l\u00e6setid":"3 minutter"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"NewsArticle","@id":"https:\/\/dailyai.com\/2024\/04\/microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm\/#article","isPartOf":{"@id":"https:\/\/dailyai.com\/2024\/04\/microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm\/"},"author":{"name":"Eugene van der Watt","@id":"https:\/\/dailyai.com\/#\/schema\/person\/7ce525c6d0c79838b7cc7cde96993cfa"},"headline":"Microsoft launches Phi-3 Mini, a tiny but powerful LM","datePublished":"2024-04-24T11:04:46+00:00","dateModified":"2024-04-24T11:07:18+00:00","mainEntityOfPage":{"@id":"https:\/\/dailyai.com\/2024\/04\/microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm\/"},"wordCount":652,"publisher":{"@id":"https:\/\/dailyai.com\/#organization"},"image":{"@id":"https:\/\/dailyai.com\/2024\/04\/microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm\/#primaryimage"},"thumbnailUrl":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/tiny-AI.webp","keywords":["LLMS","Microsoft"],"articleSection":["Industry"],"inLanguage":"da-DK"},{"@type":"WebPage","@id":"https:\/\/dailyai.com\/2024\/04\/microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm\/","url":"https:\/\/dailyai.com\/2024\/04\/microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm\/","name":"Microsoft lancerer Phi-3 Mini, en lille, men kraftfuld LM | DailyAI","isPartOf":{"@id":"https:\/\/dailyai.com\/#website"},"primaryImageOfPage":{"@id":"https:\/\/dailyai.com\/2024\/04\/microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm\/#primaryimage"},"image":{"@id":"https:\/\/dailyai.com\/2024\/04\/microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm\/#primaryimage"},"thumbnailUrl":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/tiny-AI.webp","datePublished":"2024-04-24T11:04:46+00:00","dateModified":"2024-04-24T11:07:18+00:00","breadcrumb":{"@id":"https:\/\/dailyai.com\/2024\/04\/microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm\/#breadcrumb"},"inLanguage":"da-DK","potentialAction":[{"@type":"ReadAction","target":["https:\/\/dailyai.com\/2024\/04\/microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm\/"]}]},{"@type":"ImageObject","inLanguage":"da-DK","@id":"https:\/\/dailyai.com\/2024\/04\/microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm\/#primaryimage","url":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/tiny-AI.webp","contentUrl":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/tiny-AI.webp","width":1792,"height":1024},{"@type":"BreadcrumbList","@id":"https:\/\/dailyai.com\/2024\/04\/microsoft-launches-phi-3-mini-a-tiny-but-powerful-lm\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/dailyai.com\/"},{"@type":"ListItem","position":2,"name":"Microsoft launches Phi-3 Mini, a tiny but powerful LM"}]},{"@type":"WebSite","@id":"https:\/\/dailyai.com\/#website","url":"https:\/\/dailyai.com\/","name":"DailyAI","description":"Din daglige dosis af AI-nyheder","publisher":{"@id":"https:\/\/dailyai.com\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/dailyai.com\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"da-DK"},{"@type":"Organization","@id":"https:\/\/dailyai.com\/#organization","name":"DailyAI","url":"https:\/\/dailyai.com\/","logo":{"@type":"ImageObject","inLanguage":"da-DK","@id":"https:\/\/dailyai.com\/#\/schema\/logo\/image\/","url":"https:\/\/dailyai.com\/wp-content\/uploads\/2023\/06\/Daily-Ai_TL_colour.png","contentUrl":"https:\/\/dailyai.com\/wp-content\/uploads\/2023\/06\/Daily-Ai_TL_colour.png","width":4501,"height":934,"caption":"DailyAI"},"image":{"@id":"https:\/\/dailyai.com\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/x.com\/DailyAIOfficial","https:\/\/www.linkedin.com\/company\/dailyaiofficial\/","https:\/\/www.youtube.com\/@DailyAIOfficial"]},{"@type":"Person","@id":"https:\/\/dailyai.com\/#\/schema\/person\/7ce525c6d0c79838b7cc7cde96993cfa","name":"Eugene van der Watt","image":{"@type":"ImageObject","inLanguage":"da-DK","@id":"https:\/\/dailyai.com\/wp-content\/uploads\/2023\/07\/Eugine_Profile_Picture-96x96.png","url":"https:\/\/dailyai.com\/wp-content\/uploads\/2023\/07\/Eugine_Profile_Picture-96x96.png","contentUrl":"https:\/\/dailyai.com\/wp-content\/uploads\/2023\/07\/Eugine_Profile_Picture-96x96.png","caption":"Eugene van der Watt"},"description":"Eugene har en baggrund som elektronikingeni\u00f8r og elsker alt, hvad der har med teknologi at g\u00f8re. N\u00e5r han tager en pause fra at l\u00e6se AI-nyheder, kan du finde ham ved snookerbordet.","sameAs":["www.linkedin.com\/in\/eugene-van-der-watt-16828119"],"url":"https:\/\/dailyai.com\/da\/author\/eugene\/"}]}},"_links":{"self":[{"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/posts\/11711","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/users\/6"}],"replies":[{"embeddable":true,"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/comments?post=11711"}],"version-history":[{"count":7,"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/posts\/11711\/revisions"}],"predecessor-version":[{"id":11722,"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/posts\/11711\/revisions\/11722"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/media\/11715"}],"wp:attachment":[{"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/media?parent=11711"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/categories?post=11711"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/tags?post=11711"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}