{"id":235031,"date":"2023-11-10T10:05:00","date_gmt":"2023-11-10T10:05:00","guid":{"rendered":"https:\/\/datascientest.com\/?p=235031"},"modified":"2024-04-18T14:19:46","modified_gmt":"2024-04-18T14:19:46","slug":"formation-llmops","status":"publish","type":"post","link":"https:\/\/datascientest.com\/formation-llmops","title":{"rendered":"Formation LLMOps : devenez expert en applications IA bas\u00e9es sur les LLM"},"content":{"rendered":"\t\t<div data-elementor-type=\"wp-post\" data-elementor-id=\"235031\" class=\"elementor elementor-235031\" data-elementor-post-type=\"post\">\n\t\t\t\t\t\t\t<div class=\"elementor-element elementor-element-63106d9 e-flex e-con-boxed e-con\" data-id=\"63106d9\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-28a83a5 elementor-widget elementor-widget-heading\" data-id=\"28a83a5\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<style>\/*! elementor - v3.15.0 - 02-08-2023 *\/\n.elementor-heading-title{padding:0;margin:0;line-height:1}.elementor-widget-heading .elementor-heading-title[class*=elementor-size-]>a{color:inherit;font-size:inherit;line-height:inherit}.elementor-widget-heading .elementor-heading-title.elementor-size-small{font-size:15px}.elementor-widget-heading .elementor-heading-title.elementor-size-medium{font-size:19px}.elementor-widget-heading .elementor-heading-title.elementor-size-large{font-size:29px}.elementor-widget-heading .elementor-heading-title.elementor-size-xl{font-size:39px}.elementor-widget-heading .elementor-heading-title.elementor-size-xxl{font-size:59px}<\/style><h2 class=\"elementor-heading-title elementor-size-default\">Une formation LLMOps vous permettra d\u2019acqu\u00e9rir toutes les comp\u00e9tences requises pour construire des applications bas\u00e9es sur les mod\u00e8les d\u2019intelligence artificielle de type LLM (Larges Mod\u00e8les de Langage) et \u00e0 les mettre en production ! D\u00e9couvrez tout ce qu\u2019il faut savoir\u2026<\/h2>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-d90a052 elementor-widget elementor-widget-text-editor\" data-id=\"d90a052\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<style>\/*! elementor - v3.15.0 - 02-08-2023 *\/\n.elementor-widget-text-editor.elementor-drop-cap-view-stacked .elementor-drop-cap{background-color:#69727d;color:#fff}.elementor-widget-text-editor.elementor-drop-cap-view-framed .elementor-drop-cap{color:#69727d;border:3px solid;background-color:transparent}.elementor-widget-text-editor:not(.elementor-drop-cap-view-default) .elementor-drop-cap{margin-top:8px}.elementor-widget-text-editor:not(.elementor-drop-cap-view-default) .elementor-drop-cap-letter{width:1em;height:1em}.elementor-widget-text-editor .elementor-drop-cap{float:left;text-align:center;line-height:1;font-size:50px}.elementor-widget-text-editor .elementor-drop-cap-letter{display:inline-block}<\/style>\t\t\t\t<p><span style=\"font-weight: 400;\">Depuis le lancement de <a href=\"https:\/\/datascientest.com\/meta-se-bat-a-armes-egales-contre-openai-et-gpt-4\">ChatGPT par OpenAI<\/a> fin 2022, les <a href=\"https:\/\/datascientest.com\/large-language-models-tout-savoir\">Larges Mod\u00e8les de Langage (LLM)<\/a> tels que <a href=\"https:\/\/datascientest.com\/openai-lance-gpt-4-les-5-choses-a-savoir-sur-le-successeur-de-chatgpt\">GPT-4<\/a>, <a href=\"https:\/\/datascientest.com\/claude-3-danthropic-le-llm-qui-detrone-gpt-4\">Anthropic Claude<\/a>, <a href=\"https:\/\/datascientest.com\/google-gemini-tout-savoir\">Google Gemini<\/a> ou le Fran\u00e7ais <a href=\"https:\/\/datascientest.com\/mistral-ai-devoile-le-chat-un-chatgpt-a-la-francaise\">Mistral Large<\/a> sont sous le feu des projecteurs.<\/span><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">Ces puissants <\/span><strong style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">mod\u00e8les de Deep Learning<\/strong><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">, entra\u00een\u00e9s sur des milliards de mots, peuvent g\u00e9n\u00e9rer du langage naturel \u00e0 partir d\u2019un prompt tels qu\u2019une question ou une requ\u00eate formul\u00e9e \u00e0 l\u2019\u00e9crit.<\/span><\/p><p><span style=\"font-weight: 400;\">Ils transforment la fa\u00e7on dont nous construisons et maintenons des <strong>produits bas\u00e9s sur l\u2019IA<\/strong>. Or, les <strong>applications bas\u00e9es sur les LLM<\/strong> requi\u00e8rent l\u2019\u00e9mergence de nouveaux outils et pratiques pour la gestion.<\/span><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">Par cons\u00e9quent, un nouveau terme prend de l\u2019ampleur au sein de la communaut\u00e9 du Machine Learning : le <\/span><strong style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">LLMOps<\/strong><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">.<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-d8d829c elementor-align-center button-principale elementor-widget elementor-widget-button\" data-id=\"d8d829c\" data-element_type=\"widget\" data-widget_type=\"button.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t<div class=\"elementor-button-wrapper\">\n\t\t\t<a class=\"elementor-button elementor-button-link elementor-size-sm elementor-animation-grow\" href=\"#elementor-action%3Aaction%3Dpopup%3Aopen%26settings%3DeyJpZCI6IjIzODI2MiIsInRvZ2dsZSI6ZmFsc2V9\">\n\t\t\t\t\t\t<span class=\"elementor-button-content-wrapper\">\n\t\t\t\t\t\t<span class=\"elementor-button-text\">Se former au LLMOps<\/span>\n\t\t<\/span>\n\t\t\t\t\t<\/a>\n\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-b5f20a2 e-flex e-con-boxed e-con\" data-id=\"b5f20a2\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-2830b76 elementor-widget elementor-widget-heading\" data-id=\"2830b76\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">LLMOps : qu\u2019est-ce que c\u2019est ?<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-1a84974 elementor-widget elementor-widget-text-editor\" data-id=\"1a84974\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Le terme LLMOps est l\u2019acronyme de \u00ab Large Language Model Operations \u00bb. Il s\u2019agit d\u2019un nouvel ensemble d\u2019outils et de techniques pour la <strong>gestion du cycle de vie des applications bas\u00e9es sur les LLM<\/strong>. Ceci englobe le d\u00e9veloppement, le d\u00e9ploiement et la maintenance.<\/span><\/p><p><span style=\"font-weight: 400;\">On peut aussi d\u00e9finir le LLMOps comme <strong>l\u2019\u00e9quivalent pour les LLM du<\/strong> <a href=\"https:\/\/datascientest.com\/mlops-le-devops-applique-aux-projets-de-machine-learning\">MLOps<\/a>, \u00e0 savoir les outils et meilleures pratiques pour g\u00e9rer le cycle de vie des applications bas\u00e9es sur le Machine Learning.<\/span><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">Alors, comment expliquer l\u2019<\/span><strong style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">essor du LLMOps<\/strong><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\"> ? Les premiers LLM comme <\/span><a style=\"font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align); background-color: #ffffff;\" href=\"https:\/\/datascientest.com\/bert-un-outil-de-traitement-du-langage-innovant\">BERT<\/a><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\"> et GPT-2 sont apparus aux alentours de 2018.<\/span><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">Toutefois, c\u2019est le lancement de ChatGPT en d\u00e9cembre 2022 qui a projet\u00e9 ce type de mod\u00e8les sur le devant de la sc\u00e8ne.<\/span><\/p><p><span style=\"font-weight: 400;\">Depuis lors, de nombreux chatbots similaires ont vu le jour comme Claude ou Google Gemini. Les LLM sont aussi utilis\u00e9s par les <strong>g\u00e9n\u00e9rateurs d\u2019images<\/strong> tels que <a href=\"https:\/\/datascientest.com\/dall-e-etiquette-ses-creations-le-debut-dune-ere-de-confiance\">DALL-E<\/a> et <a href=\"https:\/\/datascientest.com\/midjourney-tout-savoir\">MidJourney<\/a>, ou les g\u00e9n\u00e9rateurs de code informatique tels que <a href=\"https:\/\/datascientest.com\/gpt-4-sintegre-dans-github-copilot-tout-savoir-sur-ces-changements\">GitHub Copilot<\/a>.<\/span><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">Or, s\u2019il est tr\u00e8s facile de <\/span><strong style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">cr\u00e9er des applications<\/strong><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\"> avec les LLM, il est beaucoup plus difficile de les mettre en production. Ceci implique de nouveaux d\u00e9fis par rapport \u00e0 la <\/span><strong style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">cr\u00e9ation de produits IA<\/strong><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\"> \u00e0 l\u2019aide de mod\u00e8les de Machine Learning plus classiques.<\/span><\/p><p><span style=\"font-weight: 400;\">Afin de relever ces challenges, il est n\u00e9cessaire de <strong>d\u00e9velopper de nouveaux outils, de nouvelles techniques<\/strong> pour g\u00e9rer le cycle de vie des applications. Voil\u00e0 pourquoi le LLMOps est de plus en plus adopt\u00e9 !<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-5af06c6 elementor-widget elementor-widget-image\" data-id=\"5af06c6\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<style>\/*! elementor - v3.15.0 - 02-08-2023 *\/\n.elementor-widget-image{text-align:center}.elementor-widget-image a{display:inline-block}.elementor-widget-image a img[src$=\".svg\"]{width:48px}.elementor-widget-image img{vertical-align:middle;display:inline-block}<\/style>\t\t\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"875\" height=\"500\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/formation_llmops1.jpg\" class=\"attachment-full size-full wp-image-235076\" alt=\"\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/formation_llmops1.jpg 875w, https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/formation_llmops1-300x171.jpg 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/formation_llmops1-768x439.jpg 768w\" sizes=\"(max-width: 875px) 100vw, 875px\" \/>\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-47c4468 e-flex e-con-boxed e-con\" data-id=\"47c4468\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-7c146d6 elementor-widget elementor-widget-heading\" data-id=\"7c146d6\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">De la s\u00e9lection du mod\u00e8le au d\u00e9ploiement : les \u00e9tapes du LLMOps<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-43c164e elementor-widget elementor-widget-text-editor\" data-id=\"43c164e\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Sur certains points, les \u00e9tapes du LLMOps s\u2019apparentent \u00e0 celles du <a href=\"https:\/\/datascientest.com\/mlops-le-devops-applique-aux-projets-de-machine-learning\">MLOps<\/a>. Toutefois, la diff\u00e9rence est li\u00e9e \u00e0 la mani\u00e8re dont les <strong>applications bas\u00e9es sur les LLM<\/strong> sont construites.<\/span><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">Plut\u00f4t que d\u2019entra\u00eener les LLM en partant de z\u00e9ro, on se focalise sur l\u2019adaptation de <\/span><strong style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">LLM pr\u00e9-entra\u00een\u00e9s<\/strong><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\"> comme GPT \u00e0 des t\u00e2ches sp\u00e9cifiques. C\u2019est le \u00ab fine-tuning \u00bb<\/span><\/p><p><span style=\"font-weight: 400;\">Les mod\u00e8les de fondation sont entra\u00een\u00e9s par une poign\u00e9e d\u2019institutions dot\u00e9es de ressources informatiques massives, comme <a href=\"https:\/\/datascientest.com\/openai-lentreprise-qui-revolutionne-lia\">OpenAI<\/a> ou <a href=\"https:\/\/datascientest.com\/google-rattrape-chatgpt-avec-la-derniere-mise-a-jour-de-bard\">Google<\/a>.<\/span><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">Par exemple, selon une \u00e9tude publi\u00e9e par Lambda Labs en 2020, entra\u00eener GPT-3 avec ses 175 milliards de param\u00e8tres n\u00e9cessiterait 355 ans et 4,6 millions de dollars en utilisant une instance cloud Tesla V100.<\/span><\/p><p><span style=\"font-weight: 400;\">Ainsi, la plupart des applications sont obtenues via un l\u00e9ger r\u00e9glage d\u2019une partie du <a href=\"https:\/\/datascientest.com\/fonctionnement-des-reseaux-neurones\">r\u00e9seau de neurones<\/a>. Il peut aussi s\u2019agir d\u2019un p<strong>rocessus de Prompt Engineering (ing\u00e9nierie de prompt)<\/strong> ou d\u2019une distillation du mod\u00e8le ou des donn\u00e9es en r\u00e9seaux de moindre envergure pour des t\u00e2ches sp\u00e9cifiques.<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-0b605ac elementor-align-center button-principale elementor-widget elementor-widget-button\" data-id=\"0b605ac\" data-element_type=\"widget\" data-widget_type=\"button.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t<div class=\"elementor-button-wrapper\">\n\t\t\t<a class=\"elementor-button elementor-button-link elementor-size-sm elementor-animation-grow\" href=\"#elementor-action%3Aaction%3Dpopup%3Aopen%26settings%3DeyJpZCI6IjIzODI2MiIsInRvZ2dsZSI6ZmFsc2V9\">\n\t\t\t\t\t\t<span class=\"elementor-button-content-wrapper\">\n\t\t\t\t\t\t<span class=\"elementor-button-text\">D\u00e9couvrir le LLMOps<\/span>\n\t\t<\/span>\n\t\t\t\t\t<\/a>\n\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-4ae0309 elementor-widget elementor-widget-text-editor\" data-id=\"4ae0309\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">La premi\u00e8re \u00e9tape du LLMOps est donc la s\u00e9lection d\u2019un <strong>mod\u00e8le de fondation pr\u00e9-entra\u00een\u00e9 sur de vastes volumes de donn\u00e9es<\/strong>. Ce choix s\u2019effectue selon les crit\u00e8res de performance, de co\u00fbt, de facilit\u00e9 d\u2019utilisation et de flexibilit\u00e9.<\/span><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">On distingue deux types de mod\u00e8les. D\u2019un c\u00f4t\u00e9, les mod\u00e8les propri\u00e9taires, souvent plus larges et plus performants, comme <\/span><strong style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">OpenAI GPT, co:here, AI21 Labs Jurassic-2 ou Anthropic Claude<\/strong><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">.<\/span><\/p><p><span style=\"font-weight: 400;\">De l\u2019autre, les mod\u00e8les open-source, plus flexibles et moins co\u00fbteux, tels que <a href=\"https:\/\/datascientest.com\/stable-diffusion-xl-tout-savoir\">Stable Diffusion<\/a>, <strong>BLOOM, Meta LLaMA ou Flan-T5 par Google<\/strong>.<\/span><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">L\u2019\u00e9tape suivante est l\u2019acc\u00e8s au LLM via son <\/span><a style=\"font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align); background-color: #ffffff;\" href=\"https:\/\/datascientest.com\/openai-met-a-jour-son-api-pour-les-developpeurs\">API<\/a><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">, et son adaptation \u00e0 la t\u00e2che via le Prompt Engineering, le fine-tuning, ou encore l\u2019injection de donn\u00e9es externes.<\/span><\/p><p><span style=\"font-weight: 400;\">Comme dans le MLOps, on passe ensuite \u00e0 <strong>l\u2019\u00e9valuation des performances du mod\u00e8le<\/strong>. En l\u2018occurrence, il s\u2019agit d\u2019effectuer un <a href=\"https:\/\/datascientest.com\/a-b-testing-tout-savoir\">testing A\/B<\/a> pour mesurer la qualit\u00e9 des r\u00e9ponses produites.<\/span><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">Le mod\u00e8le peut ensuite \u00eatre d\u00e9ploy\u00e9, mais un monitoring des changements dans l\u2019API sous-jacente est indispensable. Pour cause, les LLM peuvent subir des changements drastiques \u00e0 chaque mise \u00e0 jour.<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-7a02b23 elementor-widget elementor-widget-image\" data-id=\"7a02b23\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"875\" height=\"500\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/formation_llmops2.jpg\" class=\"attachment-full size-full wp-image-235089\" alt=\"\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/formation_llmops2.jpg 875w, https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/formation_llmops2-300x171.jpg 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/formation_llmops2-768x439.jpg 768w\" sizes=\"(max-width: 875px) 100vw, 875px\" \/>\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-52af77a e-flex e-con-boxed e-con\" data-id=\"52af77a\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-b6531e5 elementor-widget elementor-widget-heading\" data-id=\"b6531e5\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">LLMOps vs MLOps : quelles sont les diff\u00e9rences ?<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-58a4bef elementor-widget elementor-widget-text-editor\" data-id=\"58a4bef\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">On d\u00e9nombre plusieurs diff\u00e9rences majeures entre MLOps et LLMOps. Les mod\u00e8les de Machine Learning sont tr\u00e8s gourmands en donn\u00e9es, car <strong>l\u2019entra\u00eenement d\u2019un r\u00e9seau de neurones<\/strong> exige beaucoup de donn\u00e9es annot\u00e9es.<\/span><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">Dans le cas des LLM, le nombre d\u2019\u00e9chantillons n\u00e9cessaires est moindre, mais une s\u00e9lection manuelle minutieuse s\u2019impose pour l\u2019apprentissage en zero-shot ou few-shot.<\/span><\/p><p><span style=\"font-weight: 400;\">Autre diff\u00e9rence cl\u00e9 : <strong>l\u2019exp\u00e9rimentation<\/strong>. En MLOps, il s\u2019agit toujours de suivre l\u2019architecture du mod\u00e8le ou les hyperparam\u00e8tres que l\u2019on entra\u00eene un mod\u00e8le en partant de z\u00e9ro ou que l\u2019on affine un mod\u00e8le pr\u00e9-entra\u00een\u00e9. Dans le cas du LLMOps, tout d\u00e9pend du choix entre Prompt Engineering et fine-tuning.<\/span><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">En ce qui concerne l\u2019\u00e9valuation des performances, on utilise un <\/span><strong style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">ensemble de validation<\/strong><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\"> pour le MLOps tandis que la plupart des organisations s\u2019en remettent au testing A\/B pour les LLM.<\/span><\/p><p><span style=\"font-weight: 400;\">Le co\u00fbt du MLOps est g\u00e9n\u00e9ralement li\u00e9 \u00e0 la collecte de donn\u00e9es et \u00e0 l\u2019entra\u00eenement du mod\u00e8le, tandis que celui du LLMOps est li\u00e9 \u00e0 <strong>l\u2019inf\u00e9rence<\/strong>.<\/span><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">Enfin, la <\/span><strong style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">latence<\/strong><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\"> est beaucoup plus probl\u00e9matique dans le LLMOps que dans le MLOps. Pour cause, elle freine la v\u00e9locit\u00e9 de l\u2019exp\u00e9rimentation pendant le d\u00e9veloppement et nuit \u00e0 l\u2019exp\u00e9rience utilisateur en production.<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-a90bc21 elementor-align-center button-principale elementor-widget elementor-widget-button\" data-id=\"a90bc21\" data-element_type=\"widget\" data-widget_type=\"button.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t<div class=\"elementor-button-wrapper\">\n\t\t\t<a class=\"elementor-button elementor-button-link elementor-size-sm elementor-animation-grow\" href=\"#elementor-action%3Aaction%3Dpopup%3Aopen%26settings%3DeyJpZCI6IjIzODI2MiIsInRvZ2dsZSI6ZmFsc2V9\">\n\t\t\t\t\t\t<span class=\"elementor-button-content-wrapper\">\n\t\t\t\t\t\t<span class=\"elementor-button-text\">Ma\u00eetriser le LLMOps<\/span>\n\t\t<\/span>\n\t\t\t\t\t<\/a>\n\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-31540c9 e-flex e-con-boxed e-con\" data-id=\"31540c9\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-b561a75 elementor-widget elementor-widget-heading\" data-id=\"b561a75\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">Pourquoi suivre une formation LLMOps ?<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-d7ec811 elementor-widget elementor-widget-text-editor\" data-id=\"d7ec811\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Qu\u2019il s\u2019agisse de chatbots intelligents, de g\u00e9n\u00e9rateurs d\u2019images, de vid\u00e9os, de code informatique, ou bien d\u2019autres types d\u2019outils d\u00e9j\u00e0 existants ou \u00e0 venir, <strong>les LLM seront de plus en plus utilis\u00e9s pour cr\u00e9er des applications.<\/strong><\/span><\/p><p><span style=\"font-weight: 400;\">Il est donc tr\u00e8s important de ma\u00eetriser les pratiques, les techniques et les outils permettant la gestion op\u00e9rationnelle et la mise en production de telles applications.<\/span><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">Une <\/span><strong style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">formation LLM<\/strong><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\"> permet de g\u00e9rer et d\u2019automatiser tout le cycle de vie, en ma\u00eetrisant des techniques comme le Prompt Engineering et le fine-tuning ou les outils tels que <\/span><strong style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">Whylabs<\/strong><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">, <\/span><strong style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">HoneyHive <\/strong><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">ou <\/span><strong style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">HumanLoop<\/strong><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">.<\/span><\/p><p><span style=\"font-weight: 400;\">D\u00e9sormais, il s\u2019agit d\u2019une expertise indispensable pour les \u00e9quipes IT, les <a href=\"https:\/\/datascientest.com\/data-science-et-metaverse\">Data Scientists<\/a>, ou encore les <a href=\"https:\/\/datascientest.com\/data-engineer-tout-savoir\">ing\u00e9nieurs de donn\u00e9es<\/a>. Elle leur permet de d\u00e9ployer, de surveiller et de maintenir efficacement les LLM et leurs applications !<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-d0d89d4 elementor-widget elementor-widget-image\" data-id=\"d0d89d4\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"875\" height=\"500\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/formation_llmops3.jpg\" class=\"attachment-full size-full wp-image-235091\" alt=\"\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/formation_llmops3.jpg 875w, https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/formation_llmops3-300x171.jpg 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/formation_llmops3-768x439.jpg 768w\" sizes=\"(max-width: 875px) 100vw, 875px\" \/>\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-7a7ec47 e-flex e-con-boxed e-con\" data-id=\"7a7ec47\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-9c481ba elementor-widget elementor-widget-heading\" data-id=\"9c481ba\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">Conclusion : la formation LLMOps, un imp\u00e9ratif pour la cr\u00e9ation d\u2019applications IA<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-2210f66 elementor-widget elementor-widget-text-editor\" data-id=\"2210f66\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Les LLM sont actuellement les <strong>mod\u00e8les pr\u00e9dominants dans le domaine de l\u2019IA<\/strong>, et vont continuer \u00e0 se d\u00e9velopper au fil des ann\u00e9es \u00e0 venir.<\/span><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">Par cons\u00e9quent, une formation LLMOps est indispensable pour <\/span><strong style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">travailler dans le domaine de l\u2019IA ou tirer profit de cette technologie r\u00e9volutionnaire<\/strong><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">.<\/span><\/p><p><span style=\"font-weight: 400;\">Afin d\u2019apprendre \u00e0 <strong>ma\u00eetriser les outils et techniques<\/strong> requis pour le d\u00e9veloppement d\u2019applications bas\u00e9es sur les LLM, vous pouvez <strong>choisir DataScientest<\/strong>.<\/span><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">Nos formations s\u2019effectuent enti\u00e8rement \u00e0 distance, et allient th\u00e9orie et pratique pour un apprentissage optimal ! Vous deviendrez <\/span><strong style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">expert dans toutes les \u00e9tapes du LLMOps<\/strong><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">.<\/span><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">Tous <\/span><strong style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">nos cursus sont \u00e9ligibles au<\/strong> <a style=\"font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align); background-color: #ffffff;\" href=\"https:\/\/datascientest.com\/cpf-comment-utiliser-son-compte-personnel-de-formation\">CPF<\/a><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">, et peuvent \u00eatre compl\u00e9t\u00e9s en <\/span><a style=\"font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align); background-color: #ffffff;\" href=\"https:\/\/datascientest.com\/definition-bootcamp\">BootCamp<\/a><span style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">, en continu ou en alternance. \u00c0 la fin du parcours, vous recevrez un dipl\u00f4me reconnu par l\u2019\u00c9tat. <\/span><strong style=\"color: var( --e-global-color-text ); font-family: var( --e-global-typography-text-font-family ), Sans-serif; text-align: var(--text-align);\">D\u00e9couvrez DataScientest !<\/strong><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-304fb89 elementor-align-center button-principale elementor-widget elementor-widget-button\" data-id=\"304fb89\" data-element_type=\"widget\" data-widget_type=\"button.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t<div class=\"elementor-button-wrapper\">\n\t\t\t<a class=\"elementor-button elementor-button-link elementor-size-sm elementor-animation-grow\" href=\"#elementor-action%3Aaction%3Dpopup%3Aopen%26settings%3DeyJpZCI6IjIzODI2MiIsInRvZ2dsZSI6ZmFsc2V9\">\n\t\t\t\t\t\t<span class=\"elementor-button-content-wrapper\">\n\t\t\t\t\t\t<span class=\"elementor-button-text\">Se former au LLMOps<\/span>\n\t\t<\/span>\n\t\t\t\t\t<\/a>\n\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-8be5344 elementor-widget elementor-widget-text-editor\" data-id=\"8be5344\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Vous savez tout sur la formation LLMOps. Pour plus d\u2019informations sur le m\u00eame sujet, d\u00e9couvrez <a href=\"https:\/\/datascientest.com\/advanced-prompt-engineering-tout-savoir\">notre dossier complet sur le Prompt Engineering<\/a> et <a href=\"https:\/\/datascientest.com\/chatgpt-4-tout-savoir\">notre dossier consacr\u00e9 \u00e0 ChatGPT<\/a> !<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-08ff88f elementor-widget elementor-widget-html\" data-id=\"08ff88f\" data-element_type=\"widget\" data-widget_type=\"html.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<script type=\"application\/ld+json\">\r\n{\r\n  \"@context\": \"https:\/\/schema.org\",\r\n  \"@type\": \"FAQPage\",\r\n  \"mainEntity\": [{\r\n    \"@type\": \"Question\",\r\n    \"name\": \"Qu\u2019est-ce que le LLMOps ?\",\r\n    \"acceptedAnswer\": {\r\n      \"@type\": \"Answer\",\r\n      \"text\": \"Le terme LLMOps est l\u2019acronyme de \u00ab Large Language Model Operations \u00bb. Il s\u2019agit d\u2019un nouvel ensemble d\u2019outils et de techniques pour la gestion du cycle de vie des applications bas\u00e9es sur les LLM. Ceci englobe le d\u00e9veloppement, le d\u00e9ploiement et la maintenance.\"\r\n    }\r\n  },{\r\n    \"@type\": \"Question\",\r\n    \"name\": \"LLMOps vs MLOps\",\r\n    \"acceptedAnswer\": {\r\n      \"@type\": \"Answer\",\r\n      \"text\": \"Les mod\u00e8les de Machine Learning sont tr\u00e8s gourmands en donn\u00e9es, car l\u2019entra\u00eenement d\u2019un r\u00e9seau de neurones exige beaucoup de donn\u00e9es annot\u00e9es. Dans le cas des LLM, le nombre d\u2019\u00e9chantillons n\u00e9cessaires est moindre. En MLOps, l\u2019exp\u00e9rimentation consiste \u00e0 suivre l\u2019architecture du mod\u00e8le ou les hyperparam\u00e8tres que l\u2019on entra\u00eene un mod\u00e8le en partant de z\u00e9ro ou que l\u2019on affine un mod\u00e8le pr\u00e9-entra\u00een\u00e9. Dans le cas du LLMOps, tout d\u00e9pend du choix entre Prompt Engineering et fine-tuning.\"\r\n    }\r\n  },{\r\n    \"@type\": \"Question\",\r\n    \"name\": \"Pourquoi suivre une formation LLMOps ?\",\r\n    \"acceptedAnswer\": {\r\n      \"@type\": \"Answer\",\r\n      \"text\": \"Qu\u2019il s\u2019agisse de chatbots intelligents, de g\u00e9n\u00e9rateurs d\u2019images, de vid\u00e9os, de code informatique, ou bien d\u2019autres types d\u2019outils d\u00e9j\u00e0 existants ou \u00e0 venir, les LLM seront de plus en plus utilis\u00e9s pour cr\u00e9er des applications. Il est donc tr\u00e8s important de ma\u00eetriser les pratiques, les techniques et les outils permettant la gestion op\u00e9rationnelle et la mise en production de telles applications.\"\r\n    }\r\n  }]\r\n}\r\n<\/script>\r\n\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t\t\t<\/div>\n\t\t","protected":false},"excerpt":{"rendered":"<p>Une formation LLMOps vous permettra d\u2019acqu\u00e9rir toutes les comp\u00e9tences requises pour construire des applications bas\u00e9es sur les mod\u00e8les d\u2019intelligence artificielle de type LLM (Larges Mod\u00e8les de Langage) et \u00e0 les mettre en production ! D\u00e9couvrez tout ce qu\u2019il faut savoir\u2026 Depuis le lancement de ChatGPT par OpenAI fin 2022, les Larges Mod\u00e8les de Langage (LLM) [&hellip;]<\/p>\n","protected":false},"author":77,"featured_media":235033,"comment_status":"open","ping_status":"open","sticky":false,"template":"elementor_theme","format":"standard","meta":{"content-type":"","inline_featured_image":false},"categories":[1417,2500],"tags":[],"yoast_head":"<!-- This site is optimized with the Yoast SEO Premium plugin v21.7 (Yoast SEO v21.7) - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>Formation LLMOps : devenez expert en applications IA bas\u00e9es sur les LLM<\/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:\/\/datascientest.com\/formation-llmops\" \/>\n<meta property=\"og:locale\" content=\"fr_FR\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Formation LLMOps : devenez expert en applications IA bas\u00e9es sur les LLM\" \/>\n<meta property=\"og:url\" content=\"https:\/\/datascientest.com\/formation-llmops\" \/>\n<meta property=\"og:site_name\" content=\"Formation Data Science | DataScientest.com\" \/>\n<meta property=\"article:published_time\" content=\"2023-11-10T10:05:00+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2024-04-18T14:19:46+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/formation_LLMOps.jpg\" \/>\n\t<meta property=\"og:image:width\" content=\"1200\" \/>\n\t<meta property=\"og:image:height\" content=\"686\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/jpeg\" \/>\n<meta name=\"author\" content=\"Raphael Kassel\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"\u00c9crit par\" \/>\n\t<meta name=\"twitter:data1\" content=\"Raphael Kassel\" \/>\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\":\"WebPage\",\"@id\":\"https:\/\/datascientest.com\/formation-llmops\",\"url\":\"https:\/\/datascientest.com\/formation-llmops\",\"name\":\"Formation LLMOps : devenez expert en applications IA bas\u00e9es sur les LLM\",\"isPartOf\":{\"@id\":\"https:\/\/datascientest.com\/#website\"},\"datePublished\":\"2023-11-10T10:05:00+00:00\",\"dateModified\":\"2024-04-18T14:19:46+00:00\",\"author\":{\"@id\":\"https:\/\/datascientest.com\/#\/schema\/person\/6eca16f63daaafcbc96e15549b9d4ee1\"},\"inLanguage\":\"fr-FR\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/datascientest.com\/formation-llmops\"]}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/datascientest.com\/#website\",\"url\":\"https:\/\/datascientest.com\/\",\"name\":\"Formation Data Science | DataScientest.com\",\"description\":\"Leader de la formation en Data Science\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/datascientest.com\/?s={search_term_string}\"},\"query-input\":\"required name=search_term_string\"}],\"inLanguage\":\"fr-FR\"},{\"@type\":\"Person\",\"@id\":\"https:\/\/datascientest.com\/#\/schema\/person\/6eca16f63daaafcbc96e15549b9d4ee1\",\"name\":\"Raphael Kassel\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"fr-FR\",\"@id\":\"https:\/\/datascientest.com\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/11\/raph.jpeg\",\"contentUrl\":\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/11\/raph.jpeg\",\"caption\":\"Raphael Kassel\"},\"description\":\"Raphael Kassel est le Head of Data Programs chez DataScientest. Il supervise une \u00e9quipe Data, impl\u00e9mente de nouveaux parcours comme \\\"Chef de projet IA\\\" et \\\"Data Product Manager\\\", et contribue \u00e0 la modernisation des op\u00e9rations du groupe. En collaboration \u00e9troite avec l'\u00e9quipe de d\u00e9veloppement, il adapte l'accompagnement acad\u00e9mique aux \u00e9volutions de l'\u00e9quipe data et \u00e0 l'internationalisation.\",\"sameAs\":[\"https:\/\/www.linkedin.com\/in\/raphaelkassel\/\",\"Data Science\"],\"url\":\"https:\/\/datascientest.com\/author\/raphaelkassel\"}]}<\/script>\n<!-- \/ Yoast SEO Premium plugin. -->","yoast_head_json":{"title":"Formation LLMOps : devenez expert en applications IA bas\u00e9es sur les LLM","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:\/\/datascientest.com\/formation-llmops","og_locale":"fr_FR","og_type":"article","og_title":"Formation LLMOps : devenez expert en applications IA bas\u00e9es sur les LLM","og_url":"https:\/\/datascientest.com\/formation-llmops","og_site_name":"Formation Data Science | DataScientest.com","article_published_time":"2023-11-10T10:05:00+00:00","article_modified_time":"2024-04-18T14:19:46+00:00","og_image":[{"width":1200,"height":686,"url":"https:\/\/datascientest.com\/wp-content\/uploads\/2024\/03\/formation_LLMOps.jpg","type":"image\/jpeg"}],"author":"Raphael Kassel","twitter_card":"summary_large_image","twitter_misc":{"\u00c9crit par":"Raphael Kassel","Dur\u00e9e de lecture estim\u00e9e":"7 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/datascientest.com\/formation-llmops","url":"https:\/\/datascientest.com\/formation-llmops","name":"Formation LLMOps : devenez expert en applications IA bas\u00e9es sur les LLM","isPartOf":{"@id":"https:\/\/datascientest.com\/#website"},"datePublished":"2023-11-10T10:05:00+00:00","dateModified":"2024-04-18T14:19:46+00:00","author":{"@id":"https:\/\/datascientest.com\/#\/schema\/person\/6eca16f63daaafcbc96e15549b9d4ee1"},"inLanguage":"fr-FR","potentialAction":[{"@type":"ReadAction","target":["https:\/\/datascientest.com\/formation-llmops"]}]},{"@type":"WebSite","@id":"https:\/\/datascientest.com\/#website","url":"https:\/\/datascientest.com\/","name":"Formation Data Science | DataScientest.com","description":"Leader de la formation en Data Science","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/datascientest.com\/?s={search_term_string}"},"query-input":"required name=search_term_string"}],"inLanguage":"fr-FR"},{"@type":"Person","@id":"https:\/\/datascientest.com\/#\/schema\/person\/6eca16f63daaafcbc96e15549b9d4ee1","name":"Raphael Kassel","image":{"@type":"ImageObject","inLanguage":"fr-FR","@id":"https:\/\/datascientest.com\/#\/schema\/person\/image\/","url":"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/11\/raph.jpeg","contentUrl":"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/11\/raph.jpeg","caption":"Raphael Kassel"},"description":"Raphael Kassel est le Head of Data Programs chez DataScientest. Il supervise une \u00e9quipe Data, impl\u00e9mente de nouveaux parcours comme \"Chef de projet IA\" et \"Data Product Manager\", et contribue \u00e0 la modernisation des op\u00e9rations du groupe. En collaboration \u00e9troite avec l'\u00e9quipe de d\u00e9veloppement, il adapte l'accompagnement acad\u00e9mique aux \u00e9volutions de l'\u00e9quipe data et \u00e0 l'internationalisation.","sameAs":["https:\/\/www.linkedin.com\/in\/raphaelkassel\/","Data Science"],"url":"https:\/\/datascientest.com\/author\/raphaelkassel"}]}},"_links":{"self":[{"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/posts\/235031"}],"collection":[{"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/users\/77"}],"replies":[{"embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/comments?post=235031"}],"version-history":[{"count":0,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/posts\/235031\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/media\/235033"}],"wp:attachment":[{"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/media?parent=235031"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/categories?post=235031"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/tags?post=235031"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}