{"id":12933,"date":"2024-08-04T16:42:03","date_gmt":"2024-08-04T21:42:03","guid":{"rendered":"http:\/\/skimai.com\/?p=12933"},"modified":"2024-08-21T18:22:21","modified_gmt":"2024-08-21T23:22:21","slug":"comment-metas-llama-3-1-repousse-les-limites-de-linformatique-libre","status":"publish","type":"post","link":"https:\/\/skimai.com\/fr\/how-metas-llama-3-1-is-pushing-the-boundaries-of-open-source-ai\/","title":{"rendered":"Le lama de Meta 3.1 : Repousser les limites de l'IA Open-Source"},"content":{"rendered":"<p>Meta a r\u00e9cemment annonc\u00e9 <a rel=\"noopener noreferrer\" href=\"https:\/\/llama.meta.com\/\">Llama 3.1<\/a>La Commission europ\u00e9enne a publi\u00e9 le premier mod\u00e8le de langage de grande taille (LLM) \u00e0 code source ouvert, le plus avanc\u00e9 \u00e0 ce jour. Cette version marque une \u00e9tape importante dans la d\u00e9mocratisation de la technologie de l'IA, en comblant potentiellement le foss\u00e9 entre les mod\u00e8les open-source et les mod\u00e8les propri\u00e9taires.<\/p>\n\n\n<p>Llama 3.1 est un grand pas en avant dans les capacit\u00e9s d'IA open-source. Avec son mod\u00e8le phare de 405 milliards de param\u00e8tres, Meta remet en question l'id\u00e9e selon laquelle l'IA de pointe doit \u00eatre ferm\u00e9e et propri\u00e9taire. Cette version marque le d\u00e9but d'une nouvelle \u00e8re o\u00f9 les capacit\u00e9s d'IA de pointe sont accessibles aux chercheurs, aux d\u00e9veloppeurs et aux entreprises de toutes tailles.<\/p>\n\n\n<p>Les principales am\u00e9liorations de Llama 3.1 comprennent une longueur de contexte \u00e9tendue \u00e0 128 000 tokens, la prise en charge de huit langues et des performances in\u00e9gal\u00e9es dans des domaines tels que le raisonnement, les math\u00e9matiques et la g\u00e9n\u00e9ration de code. Ces avanc\u00e9es font de Llama 3.1 un outil polyvalent capable de s'attaquer \u00e0 des t\u00e2ches complexes et r\u00e9elles dans divers domaines au sein de l'entreprise.<\/p>\n\n\n<div id=\"ez-toc-container\" class=\"ez-toc-v2_0_82_1 counter-hierarchy ez-toc-counter ez-toc-grey ez-toc-container-direction\">\n<div class=\"ez-toc-title-container\">\n<p class=\"ez-toc-title\" style=\"cursor:inherit\">Table des mati\u00e8res<\/p>\n<span class=\"ez-toc-title-toggle\"><a href=\"#\" class=\"ez-toc-pull-right ez-toc-btn ez-toc-btn-xs ez-toc-btn-default ez-toc-toggle\" aria-label=\"Toggle Table des mati\u00e8res\"><span class=\"ez-toc-js-icon-con\"><span class=\"\"><span class=\"eztoc-hide\" style=\"display:none;\">Toggle<\/span><span class=\"ez-toc-icon-toggle-span\"><svg style=\"fill: #999;color:#999\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" class=\"list-377408\" width=\"20px\" height=\"20px\" viewbox=\"0 0 24 24\" fill=\"none\"><path d=\"M6 6H4v2h2V6zm14 0H8v2h12V6zM4 11h2v2H4v-2zm16 0H8v2h12v-2zM4 16h2v2H4v-2zm16 0H8v2h12v-2z\" fill=\"currentColor\"><\/path><\/svg><svg style=\"fill: #999;color:#999\" class=\"arrow-unsorted-368013\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" width=\"10px\" height=\"10px\" viewbox=\"0 0 24 24\" version=\"1.2\" baseprofile=\"tiny\"><path d=\"M18.2 9.3l-6.2-6.3-6.2 6.3c-.2.2-.3.4-.3.7s.1.5.3.7c.2.2.4.3.7.3h11c.3 0 .5-.1.7-.3.2-.2.3-.5.3-.7s-.1-.5-.3-.7zM5.8 14.7l6.2 6.3 6.2-6.3c.2-.2.3-.5.3-.7s-.1-.5-.3-.7c-.2-.2-.4-.3-.7-.3h-11c-.3 0-.5.1-.7.3-.2.2-.3.5-.3.7s.1.5.3.7z\"\/><\/svg><\/span><\/span><\/span><\/a><\/span><\/div>\n<nav><ul class='ez-toc-list ez-toc-list-level-1' ><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-1\" href=\"https:\/\/skimai.com\/fr\/how-metas-llama-3-1-is-pushing-the-boundaries-of-open-source-ai\/#The_Evolution_of_Llama_From_2_to_31\" >L'\u00e9volution du lama : de 2 \u00e0 3.1<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-2\" href=\"https:\/\/skimai.com\/fr\/how-metas-llama-3-1-is-pushing-the-boundaries-of-open-source-ai\/#Technical_Specifications_of_Llama_31\" >Sp\u00e9cifications techniques de Llama 3.1<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-3\" href=\"https:\/\/skimai.com\/fr\/how-metas-llama-3-1-is-pushing-the-boundaries-of-open-source-ai\/#Breakthrough_Capabilities\" >Capacit\u00e9s de rupture<\/a><ul class='ez-toc-list-level-3' ><li class='ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-4\" href=\"https:\/\/skimai.com\/fr\/how-metas-llama-3-1-is-pushing-the-boundaries-of-open-source-ai\/#Expanded_Context_Length\" >Longueur du contexte \u00e9largi<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-5\" href=\"https:\/\/skimai.com\/fr\/how-metas-llama-3-1-is-pushing-the-boundaries-of-open-source-ai\/#Multilingual_Support\" >Support multilingue<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-6\" href=\"https:\/\/skimai.com\/fr\/how-metas-llama-3-1-is-pushing-the-boundaries-of-open-source-ai\/#Advanced_Reasoning_and_Tool_Use\" >Raisonnement avanc\u00e9 et utilisation d'outils<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-7\" href=\"https:\/\/skimai.com\/fr\/how-metas-llama-3-1-is-pushing-the-boundaries-of-open-source-ai\/#Code_Generation_and_Math_Prowess\" >G\u00e9n\u00e9ration de code et prouesses math\u00e9matiques<\/a><\/li><\/ul><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-8\" href=\"https:\/\/skimai.com\/fr\/how-metas-llama-3-1-is-pushing-the-boundaries-of-open-source-ai\/#The_Open-Source_Advantage\" >L'avantage du logiciel libre<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-9\" href=\"https:\/\/skimai.com\/fr\/how-metas-llama-3-1-is-pushing-the-boundaries-of-open-source-ai\/#Ecosystem_and_Deployment\" >Ecosyst\u00e8me et d\u00e9ploiement<\/a><ul class='ez-toc-list-level-3' ><li class='ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-10\" href=\"https:\/\/skimai.com\/fr\/how-metas-llama-3-1-is-pushing-the-boundaries-of-open-source-ai\/#Partner_Integrations\" >Int\u00e9grations de partenaires<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-11\" href=\"https:\/\/skimai.com\/fr\/how-metas-llama-3-1-is-pushing-the-boundaries-of-open-source-ai\/#Inference_Optimization_and_Scalability\" >Optimisation de l'inf\u00e9rence et \u00e9volutivit\u00e9<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-3'><a class=\"ez-toc-link ez-toc-heading-12\" href=\"https:\/\/skimai.com\/fr\/how-metas-llama-3-1-is-pushing-the-boundaries-of-open-source-ai\/#The_Llama_Stack_and_Standardization_Efforts\" >La pile de lamas et les efforts de normalisation<\/a><\/li><\/ul><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-13\" href=\"https:\/\/skimai.com\/fr\/how-metas-llama-3-1-is-pushing-the-boundaries-of-open-source-ai\/#Llama_31s_Promise_and_Potential\" >Les promesses et le potentiel de Llama 3.1<\/a><\/li><\/ul><\/nav><\/div>\n<h2 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"The_Evolution_of_Llama_From_2_to_31\"><\/span>L'\u00e9volution du lama : de 2 \u00e0 3.1<span class=\"ez-toc-section-end\"><\/span><\/h2>\n\n\n<p>Pour appr\u00e9cier l'importance de Llama 3.1, il convient de revenir sur ses pr\u00e9d\u00e9cesseurs. Llama 2, publi\u00e9 en 2023, repr\u00e9sentait d\u00e9j\u00e0 une avanc\u00e9e majeure dans le domaine de l'IA open-source. Il proposait des mod\u00e8les allant de 7B \u00e0 70B param\u00e8tres et d\u00e9montrait des performances comp\u00e9titives dans divers benchmarks.<\/p>\n\n\n<p><strong><u>Llama 3.1 s'appuie sur cette base et propose plusieurs avanc\u00e9es majeures :<\/u><\/strong><\/p>\n\n\n<ol class=\"wp-block-list\">\n<li><p><strong>Augmentation de la taille du mod\u00e8le :<\/strong> L'introduction du mod\u00e8le de param\u00e8tres 405B repousse les limites de ce qui est possible en mati\u00e8re d'IA open-source.<\/p><\/li><li><p><strong>Extension de la dur\u00e9e du contexte :<\/strong> De 4K tokens dans Llama 2 \u00e0 128K dans Llama 3.1, permettant une compr\u00e9hension plus complexe et plus nuanc\u00e9e des textes plus longs.<\/p><\/li><li><p><strong>Capacit\u00e9s multilingues :<\/strong> La prise en charge \u00e9largie des langues permet des applications plus diversifi\u00e9es dans diff\u00e9rentes r\u00e9gions et diff\u00e9rents cas d'utilisation.<\/p><\/li><li><p><strong>Am\u00e9lioration du raisonnement et des t\u00e2ches sp\u00e9cialis\u00e9es :<\/strong> Am\u00e9lioration des performances dans des domaines tels que le raisonnement math\u00e9matique et la g\u00e9n\u00e9ration de code.<\/p><\/li>\n<\/ol>\n\n\n<p>Compar\u00e9e \u00e0 des mod\u00e8les \u00e0 code source ferm\u00e9 tels que GPT-4 et Claude 3.5 Sonnet, la Llama 3.1 405B tient son rang dans divers benchmarks. Ce niveau de performance dans un mod\u00e8le \u00e0 code source ouvert est sans pr\u00e9c\u00e9dent.<\/p>\n\n\n<figure class=\"wp-block-image\">\n<img decoding=\"async\" src=\"http:\/\/skimai.com\/wp-content\/uploads\/2024\/08\/ceb48833-8f7d-4551-a3a2-8e0936a5105e.png\" alt=\"Comparaison des performances de Meta Llama 3.1\" \/>\n<\/figure>\n\n\n<h2 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"Technical_Specifications_of_Llama_31\"><\/span>Sp\u00e9cifications techniques de Llama 3.1<span class=\"ez-toc-section-end\"><\/span><\/h2>\n\n\n<p>En ce qui concerne les d\u00e9tails techniques, Llama 3.1 offre une gamme de tailles de mod\u00e8les pour r\u00e9pondre \u00e0 diff\u00e9rents besoins et ressources informatiques :<\/p>\n\n\n<ol class=\"wp-block-list\">\n<li><p><strong>Mod\u00e8le de param\u00e8tres 8B : <\/strong>Convient aux applications l\u00e9g\u00e8res et aux appareils p\u00e9riph\u00e9riques.<\/p><\/li><li><p><strong>Mod\u00e8le de param\u00e8tres 70B :<\/strong> Un \u00e9quilibre entre les exigences en mati\u00e8re de performances et de ressources.<\/p><\/li><li><p><strong>405B mod\u00e8le de param\u00e8tres :<\/strong> Le mod\u00e8le phare, qui repousse les limites des capacit\u00e9s d'IA en source ouverte.<\/p><\/li>\n<\/ol>\n\n\n<p>La m\u00e9thodologie de formation pour Llama 3.1 a impliqu\u00e9 un ensemble massif de donn\u00e9es de plus de 15 trillions de jetons, beaucoup plus important que ses pr\u00e9d\u00e9cesseurs. Ces donn\u00e9es d'entra\u00eenement \u00e9tendues, combin\u00e9es \u00e0 des techniques raffin\u00e9es de curation et de pr\u00e9traitement des donn\u00e9es, contribuent \u00e0 l'am\u00e9lioration des performances et de la polyvalence du mod\u00e8le.<\/p>\n\n\n<p>Sur le plan architectural, Llama 3.1 conserve un mod\u00e8le de transformateur de d\u00e9codeur uniquement, privil\u00e9giant la stabilit\u00e9 de l'apprentissage \u00e0 des approches plus exp\u00e9rimentales telles que le m\u00e9lange d'experts. Cependant, Meta a mis en \u0153uvre plusieurs optimisations pour permettre une formation et une inf\u00e9rence efficaces \u00e0 cette \u00e9chelle sans pr\u00e9c\u00e9dent :<\/p>\n\n\n<ol class=\"wp-block-list\">\n<li><p><strong>Infrastructure de formation \u00e9volutive : <\/strong>Utilisation de plus de 16 000 GPU H100 pour entra\u00eener le mod\u00e8le 405B.<\/p><\/li><li><p><strong>Proc\u00e9dure it\u00e9rative de post-entra\u00eenement : <\/strong>L'utilisation de la mise au point supervis\u00e9e et de l'optimisation des pr\u00e9f\u00e9rences directes pour am\u00e9liorer les capacit\u00e9s sp\u00e9cifiques.<\/p><\/li><li><p><strong>Techniques de quantification : <\/strong>R\u00e9duction du mod\u00e8le de 16 bits \u00e0 8 bits pour une inf\u00e9rence plus efficace, permettant un d\u00e9ploiement sur un seul n\u0153ud de serveur.<\/p><\/li>\n<\/ol>\n\n\n<p>Ces choix techniques refl\u00e8tent un \u00e9quilibre entre la n\u00e9cessit\u00e9 de repousser les limites de la taille des mod\u00e8les et celle d'assurer une utilisation pratique dans toute une s\u00e9rie de sc\u00e9narios de d\u00e9ploiement.<\/p>\n\n\n<p>En mettant ces mod\u00e8les avanc\u00e9s \u00e0 la disposition de tous, Meta ne se contente pas de partager un produit, mais fournit une plateforme d'innovation. Les sp\u00e9cifications techniques du Llama 3.1 offrent aux chercheurs et aux d\u00e9veloppeurs de nouvelles possibilit\u00e9s d'explorer des applications d'IA de pointe, acc\u00e9l\u00e9rant ainsi le rythme des progr\u00e8s de l'IA dans l'ensemble du secteur.<\/p>\n\n\n<figure class=\"wp-block-image\">\n<img decoding=\"async\" src=\"http:\/\/skimai.com\/wp-content\/uploads\/2024\/08\/5abb38bb-3e47-4c4e-a442-b87a4867748a.png\" alt=\"Meta Llama 3.1 architecture\" \/>\n<\/figure>\n\n\n<h2 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"Breakthrough_Capabilities\"><\/span>Capacit\u00e9s de rupture<span class=\"ez-toc-section-end\"><\/span><\/h2>\n\n\n<p>Llama 3.1 pr\u00e9sente plusieurs fonctionnalit\u00e9s r\u00e9volutionnaires qui le distinguent dans le paysage de l'IA :<\/p>\n\n\n<h3 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"Expanded_Context_Length\"><\/span>Longueur du contexte \u00e9largi<span class=\"ez-toc-section-end\"><\/span><\/h3>\n\n\n<p>Le passage \u00e0 une fen\u00eatre contextuelle de 128 Ko change la donne. Cette capacit\u00e9 \u00e9largie permet au Llama 3.1 de traiter et de comprendre des morceaux de texte beaucoup plus longs :<\/p>\n\n\n<ul class=\"wp-block-list\">\n<li><p>Analyse compl\u00e8te des documents<\/p><\/li><li><p>G\u00e9n\u00e9ration de contenu long format<\/p><\/li><li><p>Traitement plus nuanc\u00e9 des conversations<\/p><\/li>\n<\/ul>\n\n\n<p>Cette caract\u00e9ristique ouvre de nouvelles possibilit\u00e9s d'applications dans des domaines tels que le traitement des documents juridiques, l'analyse de la litt\u00e9rature et la r\u00e9solution de probl\u00e8mes complexes n\u00e9cessitant de retenir et de synth\u00e9tiser de grandes quantit\u00e9s d'informations.<\/p>\n\n\n<h3 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"Multilingual_Support\"><\/span>Support multilingue<span class=\"ez-toc-section-end\"><\/span><\/h3>\n\n\n<p>La prise en charge de huit langues par Llama 3.1 \u00e9largit consid\u00e9rablement ses possibilit\u00e9s d'application \u00e0 l'\u00e9chelle mondiale. Cette capacit\u00e9 multilingue :<\/p>\n\n\n<ul class=\"wp-block-list\">\n<li><p>Am\u00e9liore la communication interculturelle<\/p><\/li><li><p>Des applications d'IA plus inclusives<\/p><\/li><li><p>Soutenir les op\u00e9rations commerciales mondiales<\/p><\/li>\n<\/ul>\n\n\n<p>En \u00e9liminant les barri\u00e8res linguistiques, Llama 3.1 ouvre la voie \u00e0 des solutions d'IA plus diversifi\u00e9es et orient\u00e9es vers le monde.<\/p>\n\n\n<h3 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"Advanced_Reasoning_and_Tool_Use\"><\/span>Raisonnement avanc\u00e9 et utilisation d'outils<span class=\"ez-toc-section-end\"><\/span><\/h3>\n\n\n<p>Le mod\u00e8le d\u00e9montre des capacit\u00e9s de raisonnement sophistiqu\u00e9es et l'aptitude \u00e0 utiliser efficacement des outils externes. Ces progr\u00e8s se manifestent par :<\/p>\n\n\n<ul class=\"wp-block-list\">\n<li><p>Am\u00e9lioration de la d\u00e9duction logique et de la r\u00e9solution de probl\u00e8mes<\/p><\/li><li><p>Capacit\u00e9 accrue \u00e0 suivre des instructions complexes<\/p><\/li><li><p>Utilisation efficace des bases de connaissances externes et des API<\/p><\/li>\n<\/ul>\n\n\n<p>Ces capacit\u00e9s font de Llama 3.1 un outil puissant pour les t\u00e2ches n\u00e9cessitant des comp\u00e9tences cognitives de haut niveau, de la planification strat\u00e9gique \u00e0 l'analyse de donn\u00e9es complexes.<\/p>\n\n\n<h3 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"Code_Generation_and_Math_Prowess\"><\/span>G\u00e9n\u00e9ration de code et prouesses math\u00e9matiques<span class=\"ez-toc-section-end\"><\/span><\/h3>\n\n\n<p>Llama 3.1 fait preuve de capacit\u00e9s remarquables dans les domaines techniques :<\/p>\n\n\n<ul class=\"wp-block-list\">\n<li><p>G\u00e9n\u00e9rer un code fonctionnel de haute qualit\u00e9 dans plusieurs langages de programmation<\/p><\/li><li><p>R\u00e9soudre des probl\u00e8mes math\u00e9matiques complexes avec pr\u00e9cision<\/p><\/li><li><p>Aide \u00e0 la conception et \u00e0 l'optimisation des algorithmes<\/p><\/li>\n<\/ul>\n\n\n<p>Ces caract\u00e9ristiques font de Llama 3.1 un atout pr\u00e9cieux pour le d\u00e9veloppement de logiciels, le calcul scientifique et les applications d'ing\u00e9nierie.<\/p>\n\n\n<h2 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"The_Open-Source_Advantage\"><\/span>L'avantage du logiciel libre<span class=\"ez-toc-section-end\"><\/span><\/h2>\n\n\n<p>La nature open-source de Llama 3.1 apporte plusieurs avantages significatifs.<\/p>\n\n\n<p>En mettant gratuitement \u00e0 disposition des capacit\u00e9s d'IA d'avant-garde, Meta est :<\/p>\n\n\n<ul class=\"wp-block-list\">\n<li><p>Abaisser les barri\u00e8res \u00e0 l'entr\u00e9e pour la recherche et le d\u00e9veloppement de l'IA<\/p><\/li><li><p>Permettre aux petites organisations et aux d\u00e9veloppeurs individuels d'exploiter l'IA avanc\u00e9e<\/p><\/li><li><p>Favoriser un \u00e9cosyst\u00e8me de l'IA plus diversifi\u00e9 et plus innovant<\/p><\/li>\n<\/ul>\n\n\n<p>Cette d\u00e9mocratisation pourrait conduire \u00e0 une prolif\u00e9ration des applications de l'IA dans divers secteurs, ce qui pourrait acc\u00e9l\u00e9rer le progr\u00e8s technologique.<\/p>\n\n\n<p>La possibilit\u00e9 d'acc\u00e9der aux poids des mod\u00e8les de Llama 3.1 et de les modifier ouvre des perspectives de personnalisation sans pr\u00e9c\u00e9dent :<\/p>\n\n\n<ul class=\"wp-block-list\">\n<li><p>Adaptation sp\u00e9cifique \u00e0 un domaine pour les industries sp\u00e9cialis\u00e9es<\/p><\/li><li><p>Ajustement pour des cas d'utilisation et des ensembles de donn\u00e9es uniques<\/p><\/li><li><p>Exp\u00e9rimentation de nouvelles techniques et architectures de formation<\/p><\/li>\n<\/ul>\n\n\n<p>Cette flexibilit\u00e9 permet aux organisations d'adapter le mod\u00e8le \u00e0 leurs besoins sp\u00e9cifiques, ce qui peut d\u00e9boucher sur des solutions d'IA plus efficaces et plus efficientes.<\/p>\n\n\n<h2 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"Ecosystem_and_Deployment\"><\/span>Ecosyst\u00e8me et d\u00e9ploiement<span class=\"ez-toc-section-end\"><\/span><\/h2>\n\n\n<p>La sortie de Llama 3.1 s'accompagne d'un \u00e9cosyst\u00e8me solide pour soutenir son d\u00e9ploiement et son utilisation :<\/p>\n\n\n<h3 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"Partner_Integrations\"><\/span>Int\u00e9grations de partenaires<span class=\"ez-toc-section-end\"><\/span><\/h3>\n\n\n<p>Meta a collabor\u00e9 avec les leaders de l'industrie afin d'assurer un support \u00e9tendu pour Llama 3.1 :<\/p>\n\n\n<ul class=\"wp-block-list\">\n<li><p>Les fournisseurs de services en nuage tels que AWS, Google Cloud et Azure offrent des options de d\u00e9ploiement transparentes.<\/p><\/li><li><p>Les fabricants de mat\u00e9riel tels que NVIDIA et Dell fournissent une infrastructure optimis\u00e9e.<\/p><\/li><li><p>Les plateformes de donn\u00e9es telles que Databricks et Snowflake permettent un traitement efficace des donn\u00e9es et l'int\u00e9gration des mod\u00e8les.<\/p><\/li>\n<\/ul>\n\n\n<p>Ces partenariats permettent aux organisations de tirer parti de Llama 3.1 dans le cadre de leurs technologies existantes.<\/p>\n\n\n<figure class=\"wp-block-image\">\n<img decoding=\"async\" src=\"http:\/\/skimai.com\/wp-content\/uploads\/2024\/08\/f65464b4-ea93-4bca-ae3d-c4f0bb54d5a3.png\" alt=\"Caract\u00e9ristiques de Meta Llama 3.1\" \/>\n<\/figure>\n\n\n<h3 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"Inference_Optimization_and_Scalability\"><\/span>Optimisation de l'inf\u00e9rence et \u00e9volutivit\u00e9<span class=\"ez-toc-section-end\"><\/span><\/h3>\n\n\n<p>Afin de rendre Llama 3.1 pratique pour les applications du monde r\u00e9el, plusieurs optimisations ont \u00e9t\u00e9 mises en \u0153uvre :<\/p>\n\n\n<ul class=\"wp-block-list\">\n<li><p>Les techniques de quantification r\u00e9duisent les besoins de calcul du mod\u00e8le.<\/p><\/li><li><p>Des moteurs d'inf\u00e9rence optimis\u00e9s tels que vLLM et TensorRT am\u00e9liorent les performances.<\/p><\/li><li><p>Les options de d\u00e9ploiement \u00e9volutives r\u00e9pondent \u00e0 diff\u00e9rents cas d'utilisation, des appareils p\u00e9riph\u00e9riques aux centres de donn\u00e9es.<\/p><\/li>\n<\/ul>\n\n\n<p>Ces optimisations permettent de d\u00e9ployer m\u00eame le mod\u00e8le de param\u00e8tres 405B dans des environnements de production.<\/p>\n\n\n<h3 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"The_Llama_Stack_and_Standardization_Efforts\"><\/span>La pile de lamas et les efforts de normalisation<span class=\"ez-toc-section-end\"><\/span><\/h3>\n\n\n<p>Meta milite pour une normalisation de l'\u00e9cosyst\u00e8me de l'IA :<\/p>\n\n\n<ul class=\"wp-block-list\">\n<li><p>Le projet Llama Stack vise \u00e0 cr\u00e9er une interface commune pour les composants d'IA.<\/p><\/li><li><p>Des API normalis\u00e9es pourraient faciliter l'int\u00e9gration et l'interop\u00e9rabilit\u00e9 entre les diff\u00e9rents outils et plateformes d'IA.<\/p><\/li><li><p>Cette initiative pourrait d\u00e9boucher sur un \u00e9cosyst\u00e8me de d\u00e9veloppement de l'IA plus coh\u00e9rent et plus efficace.<\/p><\/li>\n<\/ul>\n\n\n<h2 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"Llama_31s_Promise_and_Potential\"><\/span>Les promesses et le potentiel de Llama 3.1<span class=\"ez-toc-section-end\"><\/span><\/h2>\n\n\n<p>La sortie de Llama 3.1 de Meta marque un tournant dans le paysage de l'IA, en d\u00e9mocratisant l'acc\u00e8s \u00e0 des capacit\u00e9s d'IA d'avant-garde. En proposant un mod\u00e8le \u00e0 405 param\u00e8tres avec des performances de pointe, un support multilingue et une longueur de contexte \u00e9tendue, le tout dans un cadre open-source, Meta a \u00e9tabli un nouveau standard pour une IA accessible et puissante. Cette initiative remet non seulement en question la domination des mod\u00e8les \u00e0 source ferm\u00e9e, mais ouvre \u00e9galement la voie \u00e0 une innovation et \u00e0 une collaboration sans pr\u00e9c\u00e9dent dans la communaut\u00e9 de l'IA. <\/p>\n\n\n<p>\u00c0 la crois\u00e9e des chemins du d\u00e9veloppement de l'intelligence artificielle, Llama 3.1 repr\u00e9sente plus qu'une simple avanc\u00e9e technologique ; il incarne la vision d'un avenir plus ouvert, plus inclusif et plus dynamique pour l'intelligence artificielle. L'impact r\u00e9el de cette version se r\u00e9v\u00e9lera lorsque les d\u00e9veloppeurs, les chercheurs et les entreprises du monde entier exploiteront son potentiel, remodelant les industries et repoussant les limites de ce qui est possible avec les LLM.<\/p>","protected":false},"excerpt":{"rendered":"<p>Meta has recently announced Llama 3.1, its most advanced open-source large language model (LLM) to date. This release marks a significant milestone in the democratization of AI technology, potentially bridging the gap between open-source and proprietary models. Llama 3.1 is a big leap forward in open-source AI capabilities. With its flagship 405 billion parameter model, [&hellip;]<\/p>\n","protected":false},"author":1003,"featured_media":12938,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"single-custom-post-template.php","format":"standard","meta":{"_et_pb_use_builder":"","_et_pb_old_content":"","_et_gb_content_width":"","footnotes":""},"categories":[125,100,67],"tags":[],"class_list":["post-12933","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-enterprise-ai-blog","category-generative-ai","category-ml-nlp"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v24.1 - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>Meta&#039;s Llama 3.1: Pushing the Boundaries of Open-Source AI - Skim AI<\/title>\n<meta name=\"description\" content=\"Explore Llama 3.1 by Meta, a groundbreaking open-source large language model with 405 billion parameters. Discover its advanced features and enterprise benefits.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/skimai.com\/fr\/comment-metas-llama-3-1-repousse-les-limites-de-linformatique-libre\/\" \/>\n<meta property=\"og:locale\" content=\"fr_FR\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Meta&#039;s Llama 3.1: Pushing the Boundaries of Open-Source AI - Skim AI\" \/>\n<meta property=\"og:description\" content=\"Explore Llama 3.1 by Meta, a groundbreaking open-source large language model with 405 billion parameters. Discover its advanced features and enterprise benefits.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/skimai.com\/fr\/comment-metas-llama-3-1-repousse-les-limites-de-linformatique-libre\/\" \/>\n<meta property=\"og:site_name\" content=\"Skim AI\" \/>\n<meta property=\"article:published_time\" content=\"2024-08-04T21:42:03+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2024-08-21T23:22:21+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/skimai.com\/wp-content\/uploads\/2024\/08\/llama-comparison.png\" \/>\n\t<meta property=\"og:image:width\" content=\"1092\" \/>\n\t<meta property=\"og:image:height\" content=\"612\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/png\" \/>\n<meta name=\"author\" content=\"Greggory Elias\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"\u00c9crit par\" \/>\n\t<meta name=\"twitter:data1\" content=\"Greggory Elias\" \/>\n\t<meta name=\"twitter:label2\" content=\"Dur\u00e9e de lecture estim\u00e9e\" \/>\n\t<meta name=\"twitter:data2\" content=\"7 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\/\/skimai.com\/uk\/\u044f\u043a-\u043c\u0435\u0442\u0430-\u043b\u0430\u043c\u0430-3-1-\u0440\u043e\u0437\u0448\u0438\u0440\u044e\u0454-\u043c\u0435\u0436\u0456-ai-\u0437-\u0432\u0456\u0434\u043a\u0440\u0438\u0442\u0438\/#article\",\"isPartOf\":{\"@id\":\"https:\/\/skimai.com\/uk\/\u044f\u043a-\u043c\u0435\u0442\u0430-\u043b\u0430\u043c\u0430-3-1-\u0440\u043e\u0437\u0448\u0438\u0440\u044e\u0454-\u043c\u0435\u0436\u0456-ai-\u0437-\u0432\u0456\u0434\u043a\u0440\u0438\u0442\u0438\/\"},\"author\":{\"name\":\"Greggory Elias\",\"@id\":\"https:\/\/skimai.com\/uk\/#\/schema\/person\/7a883b4a2d2ea22040f42a7975eb86c6\"},\"headline\":\"Meta&#8217;s Llama 3.1: Pushing the Boundaries of Open-Source AI\",\"datePublished\":\"2024-08-04T21:42:03+00:00\",\"dateModified\":\"2024-08-21T23:22:21+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\/\/skimai.com\/uk\/\u044f\u043a-\u043c\u0435\u0442\u0430-\u043b\u0430\u043c\u0430-3-1-\u0440\u043e\u0437\u0448\u0438\u0440\u044e\u0454-\u043c\u0435\u0436\u0456-ai-\u0437-\u0432\u0456\u0434\u043a\u0440\u0438\u0442\u0438\/\"},\"wordCount\":1234,\"publisher\":{\"@id\":\"https:\/\/skimai.com\/uk\/#organization\"},\"image\":{\"@id\":\"https:\/\/skimai.com\/uk\/\u044f\u043a-\u043c\u0435\u0442\u0430-\u043b\u0430\u043c\u0430-3-1-\u0440\u043e\u0437\u0448\u0438\u0440\u044e\u0454-\u043c\u0435\u0436\u0456-ai-\u0437-\u0432\u0456\u0434\u043a\u0440\u0438\u0442\u0438\/#primaryimage\"},\"thumbnailUrl\":\"https:\/\/skimai.com\/wp-content\/uploads\/2024\/08\/llama-comparison.png\",\"articleSection\":[\"Enterprise AI\",\"Generative AI\",\"LLMs \/ NLP\"],\"inLanguage\":\"fr-FR\"},{\"@type\":\"WebPage\",\"@id\":\"https:\/\/skimai.com\/uk\/\u044f\u043a-\u043c\u0435\u0442\u0430-\u043b\u0430\u043c\u0430-3-1-\u0440\u043e\u0437\u0448\u0438\u0440\u044e\u0454-\u043c\u0435\u0436\u0456-ai-\u0437-\u0432\u0456\u0434\u043a\u0440\u0438\u0442\u0438\/\",\"url\":\"https:\/\/skimai.com\/uk\/\u044f\u043a-\u043c\u0435\u0442\u0430-\u043b\u0430\u043c\u0430-3-1-\u0440\u043e\u0437\u0448\u0438\u0440\u044e\u0454-\u043c\u0435\u0436\u0456-ai-\u0437-\u0432\u0456\u0434\u043a\u0440\u0438\u0442\u0438\/\",\"name\":\"Meta's Llama 3.1: Pushing the Boundaries of Open-Source AI - Skim AI\",\"isPartOf\":{\"@id\":\"https:\/\/skimai.com\/uk\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\/\/skimai.com\/uk\/\u044f\u043a-\u043c\u0435\u0442\u0430-\u043b\u0430\u043c\u0430-3-1-\u0440\u043e\u0437\u0448\u0438\u0440\u044e\u0454-\u043c\u0435\u0436\u0456-ai-\u0437-\u0432\u0456\u0434\u043a\u0440\u0438\u0442\u0438\/#primaryimage\"},\"image\":{\"@id\":\"https:\/\/skimai.com\/uk\/\u044f\u043a-\u043c\u0435\u0442\u0430-\u043b\u0430\u043c\u0430-3-1-\u0440\u043e\u0437\u0448\u0438\u0440\u044e\u0454-\u043c\u0435\u0436\u0456-ai-\u0437-\u0432\u0456\u0434\u043a\u0440\u0438\u0442\u0438\/#primaryimage\"},\"thumbnailUrl\":\"https:\/\/skimai.com\/wp-content\/uploads\/2024\/08\/llama-comparison.png\",\"datePublished\":\"2024-08-04T21:42:03+00:00\",\"dateModified\":\"2024-08-21T23:22:21+00:00\",\"description\":\"Explore Llama 3.1 by Meta, a groundbreaking open-source large language model with 405 billion parameters. Discover its advanced features and enterprise benefits.\",\"breadcrumb\":{\"@id\":\"https:\/\/skimai.com\/uk\/\u044f\u043a-\u043c\u0435\u0442\u0430-\u043b\u0430\u043c\u0430-3-1-\u0440\u043e\u0437\u0448\u0438\u0440\u044e\u0454-\u043c\u0435\u0436\u0456-ai-\u0437-\u0432\u0456\u0434\u043a\u0440\u0438\u0442\u0438\/#breadcrumb\"},\"inLanguage\":\"fr-FR\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/skimai.com\/uk\/\u044f\u043a-\u043c\u0435\u0442\u0430-\u043b\u0430\u043c\u0430-3-1-\u0440\u043e\u0437\u0448\u0438\u0440\u044e\u0454-\u043c\u0435\u0436\u0456-ai-\u0437-\u0432\u0456\u0434\u043a\u0440\u0438\u0442\u0438\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"fr-FR\",\"@id\":\"https:\/\/skimai.com\/uk\/\u044f\u043a-\u043c\u0435\u0442\u0430-\u043b\u0430\u043c\u0430-3-1-\u0440\u043e\u0437\u0448\u0438\u0440\u044e\u0454-\u043c\u0435\u0436\u0456-ai-\u0437-\u0432\u0456\u0434\u043a\u0440\u0438\u0442\u0438\/#primaryimage\",\"url\":\"https:\/\/skimai.com\/wp-content\/uploads\/2024\/08\/llama-comparison.png\",\"contentUrl\":\"https:\/\/skimai.com\/wp-content\/uploads\/2024\/08\/llama-comparison.png\",\"width\":1092,\"height\":612,\"caption\":\"llama comparison\"},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/skimai.com\/uk\/\u044f\u043a-\u043c\u0435\u0442\u0430-\u043b\u0430\u043c\u0430-3-1-\u0440\u043e\u0437\u0448\u0438\u0440\u044e\u0454-\u043c\u0435\u0436\u0456-ai-\u0437-\u0432\u0456\u0434\u043a\u0440\u0438\u0442\u0438\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\/\/skimai.com\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Meta&#8217;s Llama 3.1: Pushing the Boundaries of Open-Source AI\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/skimai.com\/uk\/#website\",\"url\":\"https:\/\/skimai.com\/uk\/\",\"name\":\"Skim AI\",\"description\":\"The AI Agent Workforce Platform\",\"publisher\":{\"@id\":\"https:\/\/skimai.com\/uk\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/skimai.com\/uk\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"fr-FR\"},{\"@type\":\"Organization\",\"@id\":\"https:\/\/skimai.com\/uk\/#organization\",\"name\":\"Skim AI\",\"url\":\"https:\/\/skimai.com\/uk\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"fr-FR\",\"@id\":\"https:\/\/skimai.com\/uk\/#\/schema\/logo\/image\/\",\"url\":\"http:\/\/skimai.com\/wp-content\/uploads\/2020\/07\/SKIM-AI-Header-Logo.png\",\"contentUrl\":\"http:\/\/skimai.com\/wp-content\/uploads\/2020\/07\/SKIM-AI-Header-Logo.png\",\"width\":194,\"height\":58,\"caption\":\"Skim AI\"},\"image\":{\"@id\":\"https:\/\/skimai.com\/uk\/#\/schema\/logo\/image\/\"},\"sameAs\":[\"https:\/\/www.linkedin.com\/company\/skim-ai\"]},{\"@type\":\"Person\",\"@id\":\"https:\/\/skimai.com\/uk\/#\/schema\/person\/7a883b4a2d2ea22040f42a7975eb86c6\",\"name\":\"Greggory Elias\",\"url\":\"https:\/\/skimai.com\/fr\/author\/gregg\/\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Le lama de Meta 3.1 : Repousser les limites de l'IA Open-Source - Skim AI","description":"Explorez Llama 3.1 de Meta, un mod\u00e8le linguistique r\u00e9volutionnaire \u00e0 source ouverte avec 405 milliards de param\u00e8tres. D\u00e9couvrez ses fonctionnalit\u00e9s avanc\u00e9es et ses avantages pour les entreprises.","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:\/\/skimai.com\/fr\/comment-metas-llama-3-1-repousse-les-limites-de-linformatique-libre\/","og_locale":"fr_FR","og_type":"article","og_title":"Meta's Llama 3.1: Pushing the Boundaries of Open-Source AI - Skim AI","og_description":"Explore Llama 3.1 by Meta, a groundbreaking open-source large language model with 405 billion parameters. Discover its advanced features and enterprise benefits.","og_url":"https:\/\/skimai.com\/fr\/comment-metas-llama-3-1-repousse-les-limites-de-linformatique-libre\/","og_site_name":"Skim AI","article_published_time":"2024-08-04T21:42:03+00:00","article_modified_time":"2024-08-21T23:22:21+00:00","og_image":[{"width":1092,"height":612,"url":"https:\/\/skimai.com\/wp-content\/uploads\/2024\/08\/llama-comparison.png","type":"image\/png"}],"author":"Greggory Elias","twitter_card":"summary_large_image","twitter_misc":{"\u00c9crit par":"Greggory Elias","Dur\u00e9e de lecture estim\u00e9e":"7 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/skimai.com\/uk\/\u044f\u043a-\u043c\u0435\u0442\u0430-\u043b\u0430\u043c\u0430-3-1-\u0440\u043e\u0437\u0448\u0438\u0440\u044e\u0454-\u043c\u0435\u0436\u0456-ai-\u0437-\u0432\u0456\u0434\u043a\u0440\u0438\u0442\u0438\/#article","isPartOf":{"@id":"https:\/\/skimai.com\/uk\/\u044f\u043a-\u043c\u0435\u0442\u0430-\u043b\u0430\u043c\u0430-3-1-\u0440\u043e\u0437\u0448\u0438\u0440\u044e\u0454-\u043c\u0435\u0436\u0456-ai-\u0437-\u0432\u0456\u0434\u043a\u0440\u0438\u0442\u0438\/"},"author":{"name":"Greggory Elias","@id":"https:\/\/skimai.com\/uk\/#\/schema\/person\/7a883b4a2d2ea22040f42a7975eb86c6"},"headline":"Meta&#8217;s Llama 3.1: Pushing the Boundaries of Open-Source AI","datePublished":"2024-08-04T21:42:03+00:00","dateModified":"2024-08-21T23:22:21+00:00","mainEntityOfPage":{"@id":"https:\/\/skimai.com\/uk\/\u044f\u043a-\u043c\u0435\u0442\u0430-\u043b\u0430\u043c\u0430-3-1-\u0440\u043e\u0437\u0448\u0438\u0440\u044e\u0454-\u043c\u0435\u0436\u0456-ai-\u0437-\u0432\u0456\u0434\u043a\u0440\u0438\u0442\u0438\/"},"wordCount":1234,"publisher":{"@id":"https:\/\/skimai.com\/uk\/#organization"},"image":{"@id":"https:\/\/skimai.com\/uk\/\u044f\u043a-\u043c\u0435\u0442\u0430-\u043b\u0430\u043c\u0430-3-1-\u0440\u043e\u0437\u0448\u0438\u0440\u044e\u0454-\u043c\u0435\u0436\u0456-ai-\u0437-\u0432\u0456\u0434\u043a\u0440\u0438\u0442\u0438\/#primaryimage"},"thumbnailUrl":"https:\/\/skimai.com\/wp-content\/uploads\/2024\/08\/llama-comparison.png","articleSection":["Enterprise AI","Generative AI","LLMs \/ NLP"],"inLanguage":"fr-FR"},{"@type":"WebPage","@id":"https:\/\/skimai.com\/uk\/\u044f\u043a-\u043c\u0435\u0442\u0430-\u043b\u0430\u043c\u0430-3-1-\u0440\u043e\u0437\u0448\u0438\u0440\u044e\u0454-\u043c\u0435\u0436\u0456-ai-\u0437-\u0432\u0456\u0434\u043a\u0440\u0438\u0442\u0438\/","url":"https:\/\/skimai.com\/uk\/\u044f\u043a-\u043c\u0435\u0442\u0430-\u043b\u0430\u043c\u0430-3-1-\u0440\u043e\u0437\u0448\u0438\u0440\u044e\u0454-\u043c\u0435\u0436\u0456-ai-\u0437-\u0432\u0456\u0434\u043a\u0440\u0438\u0442\u0438\/","name":"Le lama de Meta 3.1 : Repousser les limites de l'IA Open-Source - Skim AI","isPartOf":{"@id":"https:\/\/skimai.com\/uk\/#website"},"primaryImageOfPage":{"@id":"https:\/\/skimai.com\/uk\/\u044f\u043a-\u043c\u0435\u0442\u0430-\u043b\u0430\u043c\u0430-3-1-\u0440\u043e\u0437\u0448\u0438\u0440\u044e\u0454-\u043c\u0435\u0436\u0456-ai-\u0437-\u0432\u0456\u0434\u043a\u0440\u0438\u0442\u0438\/#primaryimage"},"image":{"@id":"https:\/\/skimai.com\/uk\/\u044f\u043a-\u043c\u0435\u0442\u0430-\u043b\u0430\u043c\u0430-3-1-\u0440\u043e\u0437\u0448\u0438\u0440\u044e\u0454-\u043c\u0435\u0436\u0456-ai-\u0437-\u0432\u0456\u0434\u043a\u0440\u0438\u0442\u0438\/#primaryimage"},"thumbnailUrl":"https:\/\/skimai.com\/wp-content\/uploads\/2024\/08\/llama-comparison.png","datePublished":"2024-08-04T21:42:03+00:00","dateModified":"2024-08-21T23:22:21+00:00","description":"Explorez Llama 3.1 de Meta, un mod\u00e8le linguistique r\u00e9volutionnaire \u00e0 source ouverte avec 405 milliards de param\u00e8tres. D\u00e9couvrez ses fonctionnalit\u00e9s avanc\u00e9es et ses avantages pour les entreprises.","breadcrumb":{"@id":"https:\/\/skimai.com\/uk\/\u044f\u043a-\u043c\u0435\u0442\u0430-\u043b\u0430\u043c\u0430-3-1-\u0440\u043e\u0437\u0448\u0438\u0440\u044e\u0454-\u043c\u0435\u0436\u0456-ai-\u0437-\u0432\u0456\u0434\u043a\u0440\u0438\u0442\u0438\/#breadcrumb"},"inLanguage":"fr-FR","potentialAction":[{"@type":"ReadAction","target":["https:\/\/skimai.com\/uk\/\u044f\u043a-\u043c\u0435\u0442\u0430-\u043b\u0430\u043c\u0430-3-1-\u0440\u043e\u0437\u0448\u0438\u0440\u044e\u0454-\u043c\u0435\u0436\u0456-ai-\u0437-\u0432\u0456\u0434\u043a\u0440\u0438\u0442\u0438\/"]}]},{"@type":"ImageObject","inLanguage":"fr-FR","@id":"https:\/\/skimai.com\/uk\/\u044f\u043a-\u043c\u0435\u0442\u0430-\u043b\u0430\u043c\u0430-3-1-\u0440\u043e\u0437\u0448\u0438\u0440\u044e\u0454-\u043c\u0435\u0436\u0456-ai-\u0437-\u0432\u0456\u0434\u043a\u0440\u0438\u0442\u0438\/#primaryimage","url":"https:\/\/skimai.com\/wp-content\/uploads\/2024\/08\/llama-comparison.png","contentUrl":"https:\/\/skimai.com\/wp-content\/uploads\/2024\/08\/llama-comparison.png","width":1092,"height":612,"caption":"llama comparison"},{"@type":"BreadcrumbList","@id":"https:\/\/skimai.com\/uk\/\u044f\u043a-\u043c\u0435\u0442\u0430-\u043b\u0430\u043c\u0430-3-1-\u0440\u043e\u0437\u0448\u0438\u0440\u044e\u0454-\u043c\u0435\u0436\u0456-ai-\u0437-\u0432\u0456\u0434\u043a\u0440\u0438\u0442\u0438\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/skimai.com\/"},{"@type":"ListItem","position":2,"name":"Meta&#8217;s Llama 3.1: Pushing the Boundaries of Open-Source AI"}]},{"@type":"WebSite","@id":"https:\/\/skimai.com\/uk\/#website","url":"https:\/\/skimai.com\/uk\/","name":"Skim AI","description":"La plateforme de travail des agents de l'IA","publisher":{"@id":"https:\/\/skimai.com\/uk\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/skimai.com\/uk\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"fr-FR"},{"@type":"Organization","@id":"https:\/\/skimai.com\/uk\/#organization","name":"Skim AI","url":"https:\/\/skimai.com\/uk\/","logo":{"@type":"ImageObject","inLanguage":"fr-FR","@id":"https:\/\/skimai.com\/uk\/#\/schema\/logo\/image\/","url":"http:\/\/skimai.com\/wp-content\/uploads\/2020\/07\/SKIM-AI-Header-Logo.png","contentUrl":"http:\/\/skimai.com\/wp-content\/uploads\/2020\/07\/SKIM-AI-Header-Logo.png","width":194,"height":58,"caption":"Skim AI"},"image":{"@id":"https:\/\/skimai.com\/uk\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/www.linkedin.com\/company\/skim-ai"]},{"@type":"Person","@id":"https:\/\/skimai.com\/uk\/#\/schema\/person\/7a883b4a2d2ea22040f42a7975eb86c6","name":"Greggory Elias","url":"https:\/\/skimai.com\/fr\/author\/gregg\/"}]}},"_links":{"self":[{"href":"https:\/\/skimai.com\/fr\/wp-json\/wp\/v2\/posts\/12933","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/skimai.com\/fr\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/skimai.com\/fr\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/skimai.com\/fr\/wp-json\/wp\/v2\/users\/1003"}],"replies":[{"embeddable":true,"href":"https:\/\/skimai.com\/fr\/wp-json\/wp\/v2\/comments?post=12933"}],"version-history":[{"count":0,"href":"https:\/\/skimai.com\/fr\/wp-json\/wp\/v2\/posts\/12933\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/skimai.com\/fr\/wp-json\/wp\/v2\/media\/12938"}],"wp:attachment":[{"href":"https:\/\/skimai.com\/fr\/wp-json\/wp\/v2\/media?parent=12933"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/skimai.com\/fr\/wp-json\/wp\/v2\/categories?post=12933"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/skimai.com\/fr\/wp-json\/wp\/v2\/tags?post=12933"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}