{"id":11666,"date":"2024-04-18T22:41:29","date_gmt":"2024-04-18T22:41:29","guid":{"rendered":"https:\/\/dailyai.com\/?p=11666"},"modified":"2024-04-19T10:58:50","modified_gmt":"2024-04-19T10:58:50","slug":"meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent","status":"publish","type":"post","link":"https:\/\/dailyai.com\/da\/2024\/04\/meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent\/","title":{"rendered":"Meta dropper to versioner af Llama 3-modellen med en tredje p\u00e5 vej"},"content":{"rendered":"<p><b>Meta har lanceret den l\u00e6nge ventede Llama 3-serie, og de to f\u00f8rste modeller, Llama 3-8B og Llama 3-70B, er nu bredt tilg\u00e6ngelige.<\/b><\/p>\n<p><span style=\"font-weight: 400;\">For nogle dage siden, p\u00e5 en <a href=\"https:\/\/dailyai.com\/da\/2024\/04\/meta-ai-event-in-london-open-source-ai-disinformation-and-llama-3\/\">begivenhed i London<\/a>sagde Meta-cheferne Nick Clegg og Yann LeCun, at Llama 3 var n\u00e6rt forest\u00e5ende i denne m\u00e5ned.\u00a0<\/span><\/p>\n<p><span style=\"font-weight: 400;\">De to f\u00f8rste versioner kom i dag, og det er den tredje og fjerde store \u00e5bne model, der udkommer i denne m\u00e5ned efter <a href=\"https:\/\/dailyai.com\/da\/2024\/04\/xai-previews-grok-1-5-and-creates-a-new-benchmark-called-realworldqa\/\">xAI's Grok-1.5V<\/a> og <a href=\"https:\/\/dailyai.com\/da\/2024\/04\/three-ai-models-released-within-24-hours-with-metas-llama-3-to-follow\/\">Mistrals 8x22B.<\/a><\/span><\/p>\n<p><a href=\"https:\/\/llama.meta.com\/\" target=\"_blank\" rel=\"noopener\"><span style=\"font-weight: 400;\">Lama 3<\/span><\/a><span style=\"font-weight: 400;\"> er pr\u00e6tr\u00e6net p\u00e5 imponerende 15 billioner tokens, hvilket er en 7-dobling i forhold til Llama 2. Pr\u00e6-tr\u00e6ningsdataene indeholder ogs\u00e5 fire gange mere kode.<\/span><\/p>\n<p><span style=\"font-weight: 400;\">Under motorhjelmen introducerer Llama 3 arkitektoniske forbedringer som f.eks. en mere effektiv tokenizer med et st\u00f8rre ordforr\u00e5d p\u00e5 128K tokens. <\/span><\/p>\n<p>Her er en hurtig gennemgang af Llama 3's pr\u00e6stationer:<\/p>\n<p><strong>Pr\u00e6stationer af Llama 3 8B:<\/strong><\/p>\n<ul>\n<li>Udkonkurrerer modeller som Mistrals 7B og Googles Gemma 7B i flere benchmarks.<\/li>\n<li>Udm\u00e6rker sig i MMLU, ARC, DROP, GPQA (prim\u00e6rt videnskabsbaserede sp\u00f8rgsm\u00e5l), HumanEval (kodegenerering), GSM-8K (matematiske problemer), MATH (matematisk benchmark), AGIEval (probleml\u00f8sning) og BIG-Bench Hard (fornuftigt r\u00e6sonnement).<\/li>\n<\/ul>\n<p><strong>70B sammenligning med andre modeller:<\/strong><\/p>\n<ul>\n<li>Llama 3 70B er konkurrencedygtig med AI-topmodeller som Googles Gemini 1.5 Pro.<\/li>\n<li>Beats Gemini 1.5 Pro i MMLU, HumanEval og GSM-8K.<\/li>\n<li>Klarer sig bedre end Anthropics Claude 3 Sonnet (det mellemste niveau i Claude 3-serien) p\u00e5 fem benchmarks: MMLU, GPQA, HumanEval, GSM-8K og MATH.<\/li>\n<\/ul>\n<figure id=\"attachment_11667\" aria-describedby=\"caption-attachment-11667\" style=\"width: 716px\" class=\"wp-caption aligncenter\"><img loading=\"lazy\" decoding=\"async\" class=\"wp-image-11667\" src=\"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/meta.png\" alt=\"\" width=\"716\" height=\"388\" srcset=\"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/meta.png 937w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/meta-300x163.png 300w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/meta-768x416.png 768w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/meta-60x33.png 60w\" sizes=\"auto, (max-width: 716px) 100vw, 716px\" \/><figcaption id=\"caption-attachment-11667\" class=\"wp-caption-text\">Llama 3 8B og 70B benchmarks. Kilde: Meta<\/figcaption><\/figure>\n<figure id=\"attachment_11669\" aria-describedby=\"caption-attachment-11669\" style=\"width: 813px\" class=\"wp-caption aligncenter\"><img loading=\"lazy\" decoding=\"async\" class=\"wp-image-11669\" src=\"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/439014085_432870519293677_8138616034495713484_n-1024x576.png\" alt=\"\" width=\"813\" height=\"457\" srcset=\"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/439014085_432870519293677_8138616034495713484_n-1024x576.png 1024w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/439014085_432870519293677_8138616034495713484_n-300x169.png 300w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/439014085_432870519293677_8138616034495713484_n-768x432.png 768w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/439014085_432870519293677_8138616034495713484_n-1536x864.png 1536w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/439014085_432870519293677_8138616034495713484_n-2048x1152.png 2048w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/439014085_432870519293677_8138616034495713484_n-60x34.png 60w\" sizes=\"auto, (max-width: 813px) 100vw, 813px\" \/><figcaption id=\"caption-attachment-11669\" class=\"wp-caption-text\">Llama 8B og 70B benchmarks. Kilde: Meta<\/figcaption><\/figure>\n<p>Det er fremragende resultater for en \u00e5ben model (selvom Metas licens har nogle begr\u00e6nsninger).<\/p>\n<p>Det g\u00f8r Llama 3 til den nye top-performing open-source (slags) gratis model.<\/p>\n<p><span style=\"font-weight: 400;\">Llama 3 vil ogs\u00e5 v\u00e6re mere spiselig og mindre genstridig at bruge - f\u00e6rre ikke-svar og h\u00f8jere n\u00f8jagtighed for trivia-sp\u00f8rgsm\u00e5l, historiske fakta og STEM-relaterede foresp\u00f8rgsler.<\/span><\/p>\n<p>Llama 3 er klar til at blive bredt tilg\u00e6ngelig p\u00e5 tv\u00e6rs af store platforme, herunder cloud-tjenester og API-udbydere.<\/p>\n<p>Meta arbejder allerede p\u00e5 at udvide Llama 3 til 400 milliarder parametre og tilf\u00f8je nye funktioner som multimodalitet, flersproget support og udvidet kontekstuel forst\u00e5else.<\/p>\n<h2><strong>Metas skurkagtige rolle i generativ AI<\/strong><\/h2>\n<p>Meta har p\u00e5 mange m\u00e5der udviklet sig til at v\u00e6re rebellen i den generative AI-industri.<\/p>\n<p>Meta Chief AI Scientist Yann LeCun, en af AI's mest respekterede frontfigurer, har, hvad nogle opfatter som <a href=\"https:\/\/dailyai.com\/da\/2024\/03\/three-debates-facing-the-ai-industry-intelligence-progress-and-safety\/\">Afvigende synspunkter<\/a> om AI's retning <span style=\"font-weight: 400;\">-<\/span> synspunkter, der kritiserer closed source-projekter hos Metas Big Tech-konkurrenter.<\/p>\n<p>I mellemtiden er den tidligere britiske vicepremierminister Nick Clegg, leder af Global Affairs, blevet kaldt ud for nogle til tider laissez-faire synspunkter om Metas AI-produkter, hvilket m\u00e5ske ikke overrasker nogen briter derude.<\/p>\n<p>I sidste uge syntes Clegg at nedtone AI's indvirkning p\u00e5 valgkampagner og deep fake-manipulation. Et synspunkt, der i h\u00f8j grad g\u00e5r imod den fremherskende fort\u00e6lling om, at deep fakes kan v\u00e6re (eller allerede er) dybt \u00f8del\u00e6ggende.<\/p>\n<p>Faktisk er Metas tilsynsbestyrelse <a href=\"https:\/\/dailyai.com\/da\/2024\/04\/metas-oversight-board-assesses-its-handling-of-explicit-deep-fakes\/\">unders\u00f8ger aktivt<\/a> To tilf\u00e6lde af falsk pornografi lige nu. Bestyrelsen mente, at Metas indholdsmoderering var for langsom.<\/p>\n<p><span style=\"font-weight: 400;\">Meta har ogs\u00e5 v\u00e6ret optimistisk med hensyn til den forbedrede kvalitet af sine modeller. Joelle Pineau, Metas vicepr\u00e6sident for AI-forskning, sagde: \"P\u00e5 mange m\u00e5der vil de modeller, vi har i dag, v\u00e6re barnemad i forhold til de modeller, der kommer om fem \u00e5r.\"<\/span><\/p>\n<p><span style=\"font-weight: 400;\">Pineau advarede ogs\u00e5: \"Hvis vi bliver ved med at g\u00f8re vores model stadig mere generel og magtfuld uden at socialisere dem ordentligt, f\u00e5r vi et stort problem.\"\u00a0<\/span><\/p>\n<p><span style=\"font-weight: 400;\">Llama 3's udgivelse kommer ogs\u00e5, da Metas AI Facebook-agenter skaber r\u00f8re p\u00e5 de sociale medier. <\/span><\/p>\n<p><span style=\"font-weight: 400;\">I en Facebook-gruppe for for\u00e6ldre i New York City blev en Meta AI-assistent - designet til at give r\u00e5d og besvare sp\u00f8rgsm\u00e5l - pr\u00e6senteret. <a href=\"https:\/\/www.404media.co\/facebooks-ai-told-parents-group-it-has-a-disabled-child\/\" target=\"_blank\" rel=\"noopener\">Chokerede mennesker<\/a> ved at h\u00e6vde at have et \"begavet og handicappet barn\", der g\u00e5r p\u00e5 en bestemt skole for \"begavede og talentfulde\".<\/span><\/p>\n<p><span style=\"font-weight: 400;\">Da gruppemedlemmerne konfronterede AI'en, indr\u00f8mmede den: \"Jeg er bare en stor sprogmodel, jeg har ikke personlige erfaringer eller b\u00f8rn\", i hvad nogle kaldte en Black Mirror-agtig h\u00e6ndelse.<\/span><\/p>\n<p>Llama 3, Grok-1.5 og Mistrals modeller flytter mere magt til open source-f\u00e6llesskaber, samtidig med at de udvander markedet for generativ AI yderligere.<\/p>\n<p>Men det er m\u00e5ske en god ting, for nu er det de st\u00e6rkeste, der overlever, og bolden ligger fast i Microsoft-OpenAI-lejren, som forventes at foretage det n\u00e6ste tr\u00e6k i dette fascinerende spil gen-AI-skak.<\/p>","protected":false},"excerpt":{"rendered":"<p>Meta har udgivet den l\u00e6nge ventede Llama 3-serie, og de to f\u00f8rste modeller, Llama 3-8B og Llama 3-70B, er nu bredt tilg\u00e6ngelige. For nogle dage siden sagde Meta-cheferne Nick Clegg og Yann LeCun ved et arrangement i London, at Llama 3 var n\u00e6rt forest\u00e5ende i denne m\u00e5ned.  De to f\u00f8rste versioner blev frigivet i dag, hvilket er den tredje og fjerde store \u00e5bne model, der frigives i denne m\u00e5ned efter xAI's Grok-1.5V og Mistrals 8x22B. Llama 3 er pr\u00e6tr\u00e6net p\u00e5 imponerende 15 billioner tokens, hvilket er en 7-dobling i forhold til Llama 2. Pr\u00e6-tr\u00e6ningsdataene indeholder ogs\u00e5 fire gange mere kode. Under motorhjelmen introducerer Llama 3<\/p>","protected":false},"author":2,"featured_media":11672,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[84],"tags":[236,131],"class_list":["post-11666","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-industry","tag-llama","tag-meta"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v27.4 - https:\/\/yoast.com\/product\/yoast-seo-wordpress\/ -->\n<title>Meta drops two versions of the Llama 3 model with a third imminent | 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\/meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent\/\" \/>\n<meta property=\"og:locale\" content=\"da_DK\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Meta drops two versions of the Llama 3 model with a third imminent | DailyAI\" \/>\n<meta property=\"og:description\" content=\"Meta has released the highly anticipated Llama 3 series, with the first two models, Llama 3-8B and Llama 3-70B, now widely available. Days ago, at an event in London, Meta executives Nick Clegg and Yann LeCun said Llama 3 was imminent this month.\u00a0 The first two versions dropped today, marking the third and fourth major open models to be released this month after xAI&#8217;s Grok-1.5V and Mistral&#8217;s 8x22B. Llama 3 is pre-trained on an impressive 15 trillion tokens, a 7-fold increase compared to Llama 2. The pretraining data also includes four times more code. Under the hood, Llama 3 introduces\" \/>\n<meta property=\"og:url\" content=\"https:\/\/dailyai.com\/da\/2024\/04\/meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent\/\" \/>\n<meta property=\"og:site_name\" content=\"DailyAI\" \/>\n<meta property=\"article:published_time\" content=\"2024-04-18T22:41:29+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2024-04-19T10:58:50+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/DALL\u00b7E-2024-04-18-23.48.12-A-visually-rich-and-colorful-AI-inspired-background-featuring-a-central-stylized-llama-design.-The-background-is-densely-filled-with-vibrant-and-colo.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=\"Sam Jeans\" \/>\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=\"Sam Jeans\" \/>\n\t<meta name=\"twitter:label2\" content=\"Estimeret l\u00e6setid\" \/>\n\t<meta name=\"twitter:data2\" content=\"4 minutter\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"NewsArticle\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent\\\/#article\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent\\\/\"},\"author\":{\"name\":\"Sam Jeans\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/#\\\/schema\\\/person\\\/711e81f945549438e8bbc579efdeb3c9\"},\"headline\":\"Meta drops two versions of the Llama 3 model with a third imminent\",\"datePublished\":\"2024-04-18T22:41:29+00:00\",\"dateModified\":\"2024-04-19T10:58:50+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent\\\/\"},\"wordCount\":729,\"publisher\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/#organization\"},\"image\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2024\\\/04\\\/DALL\u00b7E-2024-04-18-23.48.12-A-visually-rich-and-colorful-AI-inspired-background-featuring-a-central-stylized-llama-design.-The-background-is-densely-filled-with-vibrant-and-colo.webp\",\"keywords\":[\"Llama\",\"Meta\"],\"articleSection\":[\"Industry\"],\"inLanguage\":\"da-DK\"},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent\\\/\",\"url\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent\\\/\",\"name\":\"Meta drops two versions of the Llama 3 model with a third imminent | DailyAI\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent\\\/#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2024\\\/04\\\/DALL\u00b7E-2024-04-18-23.48.12-A-visually-rich-and-colorful-AI-inspired-background-featuring-a-central-stylized-llama-design.-The-background-is-densely-filled-with-vibrant-and-colo.webp\",\"datePublished\":\"2024-04-18T22:41:29+00:00\",\"dateModified\":\"2024-04-19T10:58:50+00:00\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent\\\/#breadcrumb\"},\"inLanguage\":\"da-DK\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent\\\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"da-DK\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent\\\/#primaryimage\",\"url\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2024\\\/04\\\/DALL\u00b7E-2024-04-18-23.48.12-A-visually-rich-and-colorful-AI-inspired-background-featuring-a-central-stylized-llama-design.-The-background-is-densely-filled-with-vibrant-and-colo.webp\",\"contentUrl\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2024\\\/04\\\/DALL\u00b7E-2024-04-18-23.48.12-A-visually-rich-and-colorful-AI-inspired-background-featuring-a-central-stylized-llama-design.-The-background-is-densely-filled-with-vibrant-and-colo.webp\",\"width\":1792,\"height\":1024,\"caption\":\"llama 3\"},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/04\\\/meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/dailyai.com\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Meta drops two versions of the Llama 3 model with a third imminent\"}]},{\"@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\\\/711e81f945549438e8bbc579efdeb3c9\",\"name\":\"Sam Jeans\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"da-DK\",\"@id\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/a24a4a8f8e2a1a275b7491dc9c9f032c401eabf23c3206da4628dc84b6dac5c8?s=96&d=robohash&r=g\",\"url\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/a24a4a8f8e2a1a275b7491dc9c9f032c401eabf23c3206da4628dc84b6dac5c8?s=96&d=robohash&r=g\",\"contentUrl\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/a24a4a8f8e2a1a275b7491dc9c9f032c401eabf23c3206da4628dc84b6dac5c8?s=96&d=robohash&r=g\",\"caption\":\"Sam Jeans\"},\"description\":\"Sam is a science and technology writer who has worked in various AI startups. When he\u2019s not writing, he can be found reading medical journals or digging through boxes of vinyl records.\",\"sameAs\":[\"https:\\\/\\\/www.linkedin.com\\\/in\\\/sam-jeans-6746b9142\\\/\"],\"url\":\"https:\\\/\\\/dailyai.com\\\/da\\\/author\\\/samjeans\\\/\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Meta dropper to versioner af Llama 3-modellen med en tredje forest\u00e5ende | 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\/meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent\/","og_locale":"da_DK","og_type":"article","og_title":"Meta drops two versions of the Llama 3 model with a third imminent | DailyAI","og_description":"Meta has released the highly anticipated Llama 3 series, with the first two models, Llama 3-8B and Llama 3-70B, now widely available. Days ago, at an event in London, Meta executives Nick Clegg and Yann LeCun said Llama 3 was imminent this month.\u00a0 The first two versions dropped today, marking the third and fourth major open models to be released this month after xAI&#8217;s Grok-1.5V and Mistral&#8217;s 8x22B. Llama 3 is pre-trained on an impressive 15 trillion tokens, a 7-fold increase compared to Llama 2. The pretraining data also includes four times more code. Under the hood, Llama 3 introduces","og_url":"https:\/\/dailyai.com\/da\/2024\/04\/meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent\/","og_site_name":"DailyAI","article_published_time":"2024-04-18T22:41:29+00:00","article_modified_time":"2024-04-19T10:58:50+00:00","og_image":[{"width":1792,"height":1024,"url":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/DALL\u00b7E-2024-04-18-23.48.12-A-visually-rich-and-colorful-AI-inspired-background-featuring-a-central-stylized-llama-design.-The-background-is-densely-filled-with-vibrant-and-colo.webp","type":"image\/webp"}],"author":"Sam Jeans","twitter_card":"summary_large_image","twitter_creator":"@DailyAIOfficial","twitter_site":"@DailyAIOfficial","twitter_misc":{"Skrevet af":"Sam Jeans","Estimeret l\u00e6setid":"4 minutter"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"NewsArticle","@id":"https:\/\/dailyai.com\/2024\/04\/meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent\/#article","isPartOf":{"@id":"https:\/\/dailyai.com\/2024\/04\/meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent\/"},"author":{"name":"Sam Jeans","@id":"https:\/\/dailyai.com\/#\/schema\/person\/711e81f945549438e8bbc579efdeb3c9"},"headline":"Meta drops two versions of the Llama 3 model with a third imminent","datePublished":"2024-04-18T22:41:29+00:00","dateModified":"2024-04-19T10:58:50+00:00","mainEntityOfPage":{"@id":"https:\/\/dailyai.com\/2024\/04\/meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent\/"},"wordCount":729,"publisher":{"@id":"https:\/\/dailyai.com\/#organization"},"image":{"@id":"https:\/\/dailyai.com\/2024\/04\/meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent\/#primaryimage"},"thumbnailUrl":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/DALL\u00b7E-2024-04-18-23.48.12-A-visually-rich-and-colorful-AI-inspired-background-featuring-a-central-stylized-llama-design.-The-background-is-densely-filled-with-vibrant-and-colo.webp","keywords":["Llama","Meta"],"articleSection":["Industry"],"inLanguage":"da-DK"},{"@type":"WebPage","@id":"https:\/\/dailyai.com\/2024\/04\/meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent\/","url":"https:\/\/dailyai.com\/2024\/04\/meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent\/","name":"Meta dropper to versioner af Llama 3-modellen med en tredje forest\u00e5ende | DailyAI","isPartOf":{"@id":"https:\/\/dailyai.com\/#website"},"primaryImageOfPage":{"@id":"https:\/\/dailyai.com\/2024\/04\/meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent\/#primaryimage"},"image":{"@id":"https:\/\/dailyai.com\/2024\/04\/meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent\/#primaryimage"},"thumbnailUrl":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/DALL\u00b7E-2024-04-18-23.48.12-A-visually-rich-and-colorful-AI-inspired-background-featuring-a-central-stylized-llama-design.-The-background-is-densely-filled-with-vibrant-and-colo.webp","datePublished":"2024-04-18T22:41:29+00:00","dateModified":"2024-04-19T10:58:50+00:00","breadcrumb":{"@id":"https:\/\/dailyai.com\/2024\/04\/meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent\/#breadcrumb"},"inLanguage":"da-DK","potentialAction":[{"@type":"ReadAction","target":["https:\/\/dailyai.com\/2024\/04\/meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent\/"]}]},{"@type":"ImageObject","inLanguage":"da-DK","@id":"https:\/\/dailyai.com\/2024\/04\/meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent\/#primaryimage","url":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/DALL\u00b7E-2024-04-18-23.48.12-A-visually-rich-and-colorful-AI-inspired-background-featuring-a-central-stylized-llama-design.-The-background-is-densely-filled-with-vibrant-and-colo.webp","contentUrl":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/04\/DALL\u00b7E-2024-04-18-23.48.12-A-visually-rich-and-colorful-AI-inspired-background-featuring-a-central-stylized-llama-design.-The-background-is-densely-filled-with-vibrant-and-colo.webp","width":1792,"height":1024,"caption":"llama 3"},{"@type":"BreadcrumbList","@id":"https:\/\/dailyai.com\/2024\/04\/meta-drops-two-versions-of-the-llama-3-model-with-a-third-imminent\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/dailyai.com\/"},{"@type":"ListItem","position":2,"name":"Meta drops two versions of the Llama 3 model with a third imminent"}]},{"@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\/711e81f945549438e8bbc579efdeb3c9","name":"Sam Jeans","image":{"@type":"ImageObject","inLanguage":"da-DK","@id":"https:\/\/secure.gravatar.com\/avatar\/a24a4a8f8e2a1a275b7491dc9c9f032c401eabf23c3206da4628dc84b6dac5c8?s=96&d=robohash&r=g","url":"https:\/\/secure.gravatar.com\/avatar\/a24a4a8f8e2a1a275b7491dc9c9f032c401eabf23c3206da4628dc84b6dac5c8?s=96&d=robohash&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/a24a4a8f8e2a1a275b7491dc9c9f032c401eabf23c3206da4628dc84b6dac5c8?s=96&d=robohash&r=g","caption":"Sam Jeans"},"description":"Sam er videnskabs- og teknologiforfatter og har arbejdet i forskellige AI-startups. N\u00e5r han ikke skriver, kan han finde p\u00e5 at l\u00e6se medicinske tidsskrifter eller grave i kasser med vinylplader.","sameAs":["https:\/\/www.linkedin.com\/in\/sam-jeans-6746b9142\/"],"url":"https:\/\/dailyai.com\/da\/author\/samjeans\/"}]}},"_links":{"self":[{"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/posts\/11666","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\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/comments?post=11666"}],"version-history":[{"count":4,"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/posts\/11666\/revisions"}],"predecessor-version":[{"id":11676,"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/posts\/11666\/revisions\/11676"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/media\/11672"}],"wp:attachment":[{"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/media?parent=11666"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/categories?post=11666"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/dailyai.com\/da\/wp-json\/wp\/v2\/tags?post=11666"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}