4 étapes pour installer et exécuter Deepseek R1 localement

29 janvier 2025 White Line Ibiam Wayas

[{"selector":"#anim-5b525f7e-97bc-4cf6-ab95-93ef34fadba1","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-a8b4e28c-7ea5-43f3-95bb-7ee53acfe2dc","keyframes":{"transform":["translate3d(0px, 3340.27346%, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-993c7f04-8f70-44d3-aa02-d0ad275bc024","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-e07bc01d-3d49-419a-bf7d-935642ad1a5a","keyframes":{"transform":["translate3d(0px, 195.30265%, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] White Line Contrairement à Chatgpt, qui n'était pas ouvert à l'origine pour un usage local, l'IA profonde maintenant tendance peut fonctionner localement. Mais il y a quelques conditions préalables.

[{"selector":"#anim-f8fb3bb8-0430-474a-a352-1f68d3a67d08","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-ac21d5c9-48c5-4464-bf53-a630df3d0d57","keyframes":{"transform":["translate3d(0px, 3840.27269%, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-d859080a-59c7-4f47-a148-628ad41d5bff","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-efdfb3ea-c110-4a71-bc7a-390a60e2daad","keyframes":{"transform":["translate3d(0px, 127.95536%, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] White Line GPU: NVIDIA GPU avec au moins 12 Go de VRAM pour les modèles légers ou au moins 24 Go pour les modèles plus lourds. RAM: au moins 16 Go de mémoire système (32 Go recommandée). Espace disque: 500 Go (peut varier selon les modèles.

Étape 1: Installer Olllama

[{"selector":"#anim-c76cfe2e-9dba-4e35-a094-6ea762bd5e56","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-bb259c90-1956-4281-bdfd-42548a7e88e5","keyframes":{"transform":["translate3d(0px, 3340.27346%, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-2e3a04f2-eaaa-4bf0-8174-84fbc2cb22fb","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-47307265-9ecd-48bc-a044-c58370c5b584","keyframes":{"transform":["translate3d(0px, 316.08199%, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-7d6c684f-b872-4e65-afc2-29164a55fec3","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-43a41a45-942f-46a1-9164-cdf4feb1f8fb","keyframes":{"transform":["translate3d(0px, 981.94512%, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] White Line Olllama est simplement un outil léger pour gérer et exécuter les modèles AI localement sur votre machine.

[{"selector":"#anim-12590de8-c4b0-4e15-a28b-90b4b5b85437","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-5ef4cd2e-7575-4ea6-8c39-869d783aeab3","keyframes":{"transform":["translate3d(0px, 3340.27346%, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-94bdaa5b-75e4-4986-9457-1703d0a99357","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-da1bdcf2-506d-4436-99cf-86a75174cf4e","keyframes":{"transform":["translate3d(0px, 195.30265%, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] White Line Donc, téléchargez une version compatible d'Ollma sur le site officiel et installez-la sur votre machine, en suivant les instructions données.

Vous aimerez peut-être aussi

[{"selector":"#anim-8c6eb076-e45f-442b-bbca-62ed7b09a672","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-f82a7de1-7a55-46e9-b775-571a076fb0fc","keyframes":{"transform":["translate3d(-716.66701%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-0a45567f-8d52-494c-b2f6-26cec44ca13e","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-a4a8f4e2-c143-4329-a229-843fd78ccfc9","keyframes":{"transform":["translate3d(-130.12821%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-945d18ea-7aa4-4bd5-96b1-f15e128688ea","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-b67c3c5d-9016-47b6-9502-17aee3f27994","keyframes":{"transform":["translate3d(722.22260%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-3a3c5c46-ea29-4b60-8ce8-a24aad5d09d8","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-c6e63ef8-5821-4545-9fa8-1a2a631efbcd","keyframes":{"transform":["translate3d(132.05127%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] Deepseek est différent de tout ce que Openai a vu malgré la confiance de Sam Altman Arrow La marine de President Trump interdit officiellement l'IA profonde de la Chine en Amérique Arrow

Étape 2: Vérifiez l'installation d'Ollma

[{"selector":"#anim-1b6b28dd-adca-4fd4-8450-7ea2f1d92eb9","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-5977a981-97d4-41c8-a240-7e5964f9d979","keyframes":{"transform":["translate3d(0px, 3340.27346%, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-5d3b9daf-25b8-45ca-830b-01085d02e1fa","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-36b44df9-1953-4c28-b822-13a0ca7af77f","keyframes":{"transform":["translate3d(0px, 164.53574%, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-8688e476-ac62-4493-a378-f0c51c723cfd","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-257c4e59-82e2-4e04-9637-eef6a4355f5b","keyframes":{"transform":["translate3d(0px, 556.10204%, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] White Line A confirmé l'installation en ouvrant un nouveau terminal et en exécutant la commande «Olllama --version», cela devrait renvoyer la version d'Olllama, si elle est installée correctement.

Étape 3: Déployer le modèle R1 Deepseek

[{"selector":"#anim-35317f7c-b020-4051-825a-039ddd722867","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-5d012733-4bb0-4186-a26f-7a5e978abd2f","keyframes":{"transform":["translate3d(0px, 556.10204%, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-aee9a71f-8732-49f0-9eab-2089fbb3efc8","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-91175754-25f5-49d3-8933-d8cbe22a2a83","keyframes":{"transform":["translate3d(0px, 3340.27346%, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-c3ad7afe-39d1-4501-98e2-8d163f0e8fb8","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-b3194a75-3790-4a92-89ed-bd9f254aa4da","keyframes":{"transform":["translate3d(0px, 242.64867%, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] White Line La prochaine chose est de télécharger le modèle Deepseek R1 avec votre type de modèle préféré, comme 8b, 70b, etc.

[{"selector":"#anim-53e5e413-5201-489c-a8d8-2c142d8711cd","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-1ddae0ee-9848-4b3a-85a6-2feec2c48666","keyframes":{"transform":["translate3d(0px, 3840.27269%, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-6d7c8287-7bbc-4c43-a8a0-3d6ed500cb3c","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-5b607352-9544-48b3-9af7-3e07bc8fdf7a","keyframes":{"transform":["translate3d(0px, 227.82297%, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] White Line Vous pouvez facilement le faire sur Ollama en ouvrant votre terminal et en tapant la commande: «Olllama Run Deepseek-R1» et le code du modèle.

Étape 4: Commencez à courir en profondeur R1

[{"selector":"#anim-1d1cfde2-30f2-4aca-96d2-718fe1ec3c7a","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-ea30ecde-ac95-4bc7-8784-98371532f40a","keyframes":{"transform":["translate3d(0px, 588.88887%, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-a309b7df-a37d-4194-bdfd-d8e337594273","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-0bfe6c91-4c44-4814-a41e-86e1a99d2a48","keyframes":{"transform":["translate3d(0px, 3590.27346%, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-96f7dc0f-5dad-4782-beaf-e0aaed49a0ff","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-cb938c3c-dd19-4344-adcf-5bbd34add143","keyframes":{"transform":["translate3d(0px, 132.76645%, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] White Line La commande démarre une invite pour télécharger le modèle R1. Une fois le téléchargement terminé, Olllama ouvriramaticune console pour vous pour taper et envoyer une invite au modèle. C'est là que vous pouvez discuter avec R1 localement.

Deepseek est-il meilleur que Chatgpt?

Suivant White Line Lire maintenant