{"id":119731,"date":"2024-09-07T21:43:07","date_gmt":"2024-09-07T19:43:07","guid":{"rendered":"https:\/\/prohoster.info\/blog\/novosti-interneta\/red-hat-opublikoval-pervyj-vypusk-distributiva-red-hat-enterprise-linux-ai"},"modified":"2024-09-07T21:43:07","modified_gmt":"2024-09-07T19:43:07","slug":"red-hat-opublikoval-pervyj-vypusk-distributiva-red-hat-enterprise-linux-ai","status":"publish","type":"post","link":"https:\/\/prohoster.info\/sq\/blog\/novosti-interneta\/red-hat-opublikoval-pervyj-vypusk-distributiva-red-hat-enterprise-linux-ai","title":{"rendered":"Red Hat publikoi versionin e par\u00eb t\u00eb shp\u00ebrndarjes Red Hat Enterprise Linux AI","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"
Kompania Red Hat shpalli gatishm\u00ebrin\u00eb e versionit t\u00eb ri t\u00eb shp\u00ebrndarjes RHEL - Red Hat Enterprise Linux AI (RHEL AI), e cila \u00ebsht\u00eb p\u00ebrshtatur posa\u00e7\u00ebrisht p\u00ebr t\u00eb kryer detyra t\u00eb m\u00ebsimit t\u00eb makinerive dhe ka si q\u00ebllim t\u00eb thjeshtoj\u00eb krijimin e zgjidhjeve serverike q\u00eb p\u00ebrdorin modele gjuh\u00ebsore t\u00eb m\u00ebdha. <\/p>\n
N\u00eb p\u00ebrb\u00ebrje t\u00eb distribuimit \u00ebsht\u00eb nj\u00eb grup mjetesh dhe kornizash p\u00ebr m\u00ebsimin e makin\u00ebs, drajver\u00eb p\u00ebr p\u00ebrdorimin e akselerator\u00ebve t\u00eb ndrysh\u00ebm harduerik\u00eb nga AMD, Intel dhe NVIDIA, si dhe componente p\u00ebr aktivizimin e mund\u00ebsive t\u00eb server\u00ebve Dell, Cisco, HPE, Lenovo dhe SuperMicro, t\u00eb optimizuar p\u00ebr sistemet AI. Nd\u00ebrtesat shp\u00ebrndahen p\u00ebrmes Red Hat Customer Portal n\u00eb variante p\u00ebr instalim t\u00eb drejtp\u00ebrdrejt\u00eb n\u00eb server\u00eb dhe p\u00ebr implementim n\u00eb sistemet cloud AWS dhe IBM Cloud. <\/p>\n
Platforma mund t\u00eb p\u00ebrdoret p\u00ebr zhvillimin e aplikacioneve t\u00eb AI p\u00ebr nd\u00ebrmarrje, implementimin e sh\u00ebrbimeve t\u00eb gjenerimit t\u00eb p\u00ebrmbajtjes, krijimin e sistemeve dialoguese dhe integrimin n\u00eb aplikacionet e asistent\u00ebve virtual\u00eb, duke mb\u00ebshtetur aft\u00ebsi t\u00eb tilla si p\u00ebrgjigja n\u00eb pyetje n\u00eb gjuh\u00ebn natyrale, zgjidhja e problemeve matematikore, gjenerimi i tekstit t\u00eb kuptuesh\u00ebm mbi nj\u00eb tem\u00eb t\u00eb caktuar, p\u00ebrgatitja e p\u00ebrmbledhjeve t\u00eb p\u00ebrmbajtjes, korrigjimi i gabimeve n\u00eb tekst, kryerja e rishkrimit me fjal\u00eb t\u00eb tjera, ndihma n\u00eb shkrimin e kodit n\u00eb gjuh\u00eb t\u00eb ndryshme programimi, dhe form\u00ebsimi i letrave dhe dokumenteve n\u00eb p\u00ebrputhje me nj\u00eb model. <\/p>\n
N\u00eb RHEL AI jan\u00eb t\u00eb integruara komponent\u00ebt p\u00ebr krijimin, testimin dhe ekzekutimin e sistemeve t\u00eb m\u00ebsimit t\u00eb makinerive t\u00eb bazuara n\u00eb modelin e madh t\u00eb gjuh\u00ebs Granite, i hapur nga IBM n\u00ebn licenc\u00ebn Apache 2.0, i cili \u00ebsht\u00eb i aft\u00eb t\u00eb marr\u00eb n\u00eb konsiderat\u00eb deri n\u00eb 4,000 token gjat\u00eb gjenerimit t\u00eb tekstit dhe mbulon 7 miliard\u00eb parametra. P\u00ebr t\u00eb nd\u00ebrvepruar me modelin Granite, n\u00eb distribucion \u00ebsht\u00eb integruar mjeti i hapur InstructLab, i cili mb\u00ebshtet metodologjin\u00eb LAB (Large-scale Alignment for chatBots) p\u00ebr p\u00ebrshtatjen sipas nevojave dhe optimizimin e modeleve, si dhe p\u00ebr shtimin e njohurive shtes\u00eb dhe realizimin e aft\u00ebsive t\u00eb reja n\u00eb modelet e trajnuara paraprakisht. \n Burimi: opennet.ru<\/a> <\/p>","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"excerpt":{"rendered":"