{"id":86523,"date":"2020-06-26T07:42:14","date_gmt":"2020-06-26T05:42:14","guid":{"rendered":"https:\/\/prohoster.info\/blog\/administrirovanie\/mlops-devops-v-mire-machine-learning"},"modified":"2020-06-26T07:42:14","modified_gmt":"2020-06-26T05:42:14","slug":"mlops-devops-v-mire-machine-learning","status":"publish","type":"post","link":"https:\/\/prohoster.info\/sq\/blog\/administrirovanie\/mlops-devops-v-mire-machine-learning","title":{"rendered":"MLOps: DevOps n\u00eb bot\u00ebn e Machine Learning","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"<p>N\u00eb vitin 2018, n\u00eb qarqet profesionale dhe n\u00eb konferencat tematike q\u00eb trajtonin AI, u shfaq termi MLOps, i cili shpejt u konsolidua n\u00eb industri dhe tani po zhvillohet si nj\u00eb drejtim i pavarur. N\u00eb perspektiv\u00eb, MLOps mund t\u00eb b\u00ebhet nj\u00eb nga fushat m\u00eb t\u00eb k\u00ebrkuara n\u00eb IT. \u00c7far\u00eb \u00ebsht\u00eb kjo dhe si funksionon, e kuptojm\u00eb m\u00eb posht\u00eb.<\/p>\n<p><img decoding=\"async\" alt=\"MLOps: DevOps n\u00eb bot\u00ebn e Machine Learning\" src=\"\/wp-content\/uploads\/2020\/06\/03bcaaaf5b9ddd864c7e8e0a607d81c4.jpeg\" style=\"display:block;margin: 0 auto;\" \/><br \/>\n<noindex><a rel=\"nofollow\" name=\"habracut\"><\/a><\/noindex><\/p>\n<h2>\u00c7far\u00eb \u00ebsht\u00eb MLOps<\/h2>\n<p>\nMLOps (bashkimi i teknologjive dhe proceseve t\u00eb m\u00ebsimit t\u00eb avtomatizuar dhe qasjet p\u00ebr implementimin e modeleve t\u00eb zhvilluara n\u00eb proceset biznesore) \u00ebsht\u00eb nj\u00eb m\u00ebnyr\u00eb e re bashk\u00ebpunimi mes p\u00ebrfaq\u00ebsuesve t\u00eb biznesit, shkenc\u00ebtar\u00ebve,\u6570\u5b66ik\u00ebve, specialist\u00ebve n\u00eb fush\u00ebn e m\u00ebsimit t\u00eb automatik dhe inxhinier\u00ebve t\u00eb IT-s\u00eb p\u00ebr krijimin e sistemeve t\u00eb inteligjenc\u00ebs artificiale.<\/p>\n<p>Me fjal\u00eb t\u00eb tjera, kjo \u00ebsht\u00eb nj\u00eb m\u00ebnyr\u00eb p\u00ebr t\u00eb transformuar metodat dhe teknologjit\u00eb e m\u00ebsimit t\u00eb automatik n\u00eb nj\u00eb instrument t\u00eb dobish\u00ebm p\u00ebr zgjidhjen e problemeve t\u00eb biznesit.\u00a0<\/p>\n<p>Duhet kuptuar se zinxhiri i produktivizimit fillon shum\u00eb p\u00ebrpara zhvillimit t\u00eb modelit. Hapi i par\u00eb \u00ebsht\u00eb p\u00ebrcaktimi i problemit t\u00eb biznesit, hipoteza p\u00ebr vler\u00ebn q\u00eb mund t\u00eb nxirret nga t\u00eb dh\u00ebnat dhe ideja e biznesit p\u00ebr aplikimin e saj.\u00a0<\/p>\n<p>Vet\u00eb termi MLOps ka lindur si nj\u00eb analogji e termit DevOps n\u00eb lidhje me modelet dhe teknologjit\u00eb e m\u00ebsimit t\u00eb automatik. DevOps \u00ebsht\u00eb nj\u00eb qasje p\u00ebr zhvillimin e softuerit q\u00eb lejon t\u00eb rritet shpejt\u00ebsia e implementimit t\u00eb ndryshimeve t\u00eb ve\u00e7anta nd\u00ebrsa ruan fleksibilitetin dhe besueshm\u00ebrin\u00eb fal\u00eb nj\u00eb serie qasjesh, mes t\u00eb cilave zhvillimi i pand\u00ebrprer\u00eb, ndarja e funksioneve n\u00eb nj\u00eb s\u00ebr\u00eb mikrosherbimesh t\u00eb pavarura, testimi automatizuar dhe implementimi i ndryshimeve t\u00eb ve\u00e7anta, monitorimi global i operativitetit, sistemi p\u00ebr p\u00ebrgjigjen e shpejt\u00eb ndaj \u00e7rregullimeve t\u00eb identifikuara etj.\u00a0<\/p>\n<p>DevOps p\u00ebrcaktoi ciklin e jet\u00ebs s\u00eb softuerit dhe n\u00eb komunitetin e specialist\u00ebve u shfaq ideja p\u00ebr t\u00eb p\u00ebrdorur t\u00eb nj\u00ebjt\u00ebn metodik\u00eb n\u00eb lidhje me t\u00eb dh\u00ebnat e m\u00ebdha. DataOps \u00ebsht\u00eb nj\u00eb p\u00ebrpjekje p\u00ebr t\u00eb adaptohet dhe zgjeruar metodik\u00ebn duke marr\u00eb parasysh karakteristikat e ruajtjes, transmetimit dhe p\u00ebrpunimit t\u00eb masave t\u00eb m\u00ebdha t\u00eb t\u00eb dh\u00ebnave n\u00eb platforma t\u00eb ndryshme dhe q\u00eb nd\u00ebrveprojn\u00eb me nj\u00ebra-tjetr\u00ebn.<br \/>\n\u00a0\u00a0<br \/>\nMe shum\u00eb modele t\u00eb m\u00ebsimit t\u00eb makinerive t\u00eb implementuara n\u00eb proceset biznesore, \u00ebsht\u00eb v\u00ebn\u00eb re nj\u00eb ngjashm\u00ebri e fort\u00eb midis ciklit t\u00eb jet\u00ebs s\u00eb modeleve t\u00eb m\u00ebsimit t\u00eb makinerive dhe ciklit t\u00eb jet\u00ebs s\u00eb softuerit. Diferenca \u00ebsht\u00eb vet\u00ebm n\u00eb at\u00eb q\u00eb algoritmet e modeleve krijohen me mjete dhe metoda t\u00eb m\u00ebsimit t\u00eb makinerive. Prandaj, \u00ebsht\u00eb shfaqur natyrsh\u00ebm ideja p\u00ebr t\u00eb aplikuar dhe p\u00ebrshtatur qasjet e njohura p\u00ebr zhvillimin e softuerit p\u00ebr modele t\u00eb m\u00ebsimit t\u00eb makinerive. K\u00ebshtu, n\u00eb ciklin e jet\u00ebs s\u00eb modeleve t\u00eb m\u00ebsimit t\u00eb makinerive, mund t\u00eb ve\u00e7ojm\u00eb k\u00ebto hapa ky\u00e7:<\/p>\n<ul>\n<li>definimi i ides\u00eb s\u00eb biznesit;\n<\/li>\n<li>trajnim i modelit;\n<\/li>\n<li>testimi dhe implementimi i modelit n\u00eb procesin e biznesit;\n<\/li>\n<li>eksplorimi i modelit.\n<\/li>\n<\/ul>\n<p>\nKur gjat\u00eb p\u00ebrdorimit shfaqet nevoja p\u00ebr t\u00eb ndryshuar ose p\u00ebr t\u00eb trajnuar p\u00ebrs\u00ebri modelin mbi t\u00eb dh\u00ebna t\u00eb reja, cikli nis nga e para \u2014 modeli p\u00ebrmir\u00ebsohet, testohet dhe versioni i ri implementohet.<\/p>\n<blockquote><p>Shtes\u00eb. Pse t\u00eb trajnohet p\u00ebrs\u00ebri dhe jo t\u00eb t\u00ebrhiqet? Termi \u201ct\u00ebrheqja e modelit\u201d ka dykuptim\u00ebsi: p\u00ebr specialist\u00ebt ai do t\u00eb thot\u00eb nj\u00eb defekt t\u00eb modelit, kur modeli parashikon mir\u00eb, n\u00eb fakt p\u00ebrs\u00ebrit parametrin e parashikuar n\u00eb grupin e t\u00eb dh\u00ebnave t\u00eb trajnuara, por punon shum\u00eb m\u00eb keq n\u00eb grupin e t\u00eb dh\u00ebnave t\u00eb jashtme. Natyrisht, nj\u00eb model i till\u00eb \u00ebsht\u00eb nj\u00eb defekt, pasi ky problem nuk lejon p\u00ebrdorimin e tij.<\/p><\/blockquote>\n<p>\nN\u00eb k\u00ebt\u00eb cik\u00ebl t\u00eb jet\u00ebs, duket logjike p\u00ebrdorimi i mjeteve DevOps: testim automatizuar, implementim dhe monitorim, formalizimi i llogaritjeve t\u00eb modeleve si mikrosh\u00ebrbime t\u00eb ve\u00e7anta. Por ka edhe disa ve\u00e7ori q\u00eb pengojn\u00eb aplikimin e drejtp\u00ebrdrejt\u00eb t\u00eb k\u00ebtyre mjeteve pa nj\u00eb lidhje t\u00eb m\u00ebtejshme t\u00eb ML.<\/p>\n<p><img decoding=\"async\" alt=\"MLOps: DevOps n\u00eb bot\u00ebn e Machine Learning\" src=\"\/wp-content\/uploads\/2020\/06\/e1d01fd68c3f0071f925a82c08d29d25.jpeg\" style=\"display:block;margin: 0 auto;\" \/><br \/>\n<\/p>\n<h2>Si t'i b\u00ebjm\u00eb modelet t\u00eb funksionojn\u00eb dhe t\u00eb sjellin fitim<\/h2>\n<p>\nSi shembull, me t\u00eb cilin do t\u00eb demonstrojm\u00eb zbatimin e qasjes MLOps, do t\u00eb marrim detyr\u00ebn klasike t\u00eb robotizimit t\u00eb bisedave mb\u00ebshtet\u00ebse p\u00ebr produktet bankare (ose ndonj\u00eb produkt tjet\u00ebr). Zakonisht procesi i biznesit p\u00ebr mb\u00ebshtetje p\u00ebrmes bisedave duket si m\u00eb posht\u00eb: klienti shkruan nj\u00eb mesazh me pyetje n\u00eb bised\u00eb dhe merr p\u00ebrgjigjen e specialistit brenda nj\u00eb strukture t\u00eb caktuar dialogu. Detyra e automatizimit t\u00eb nj\u00eb bised\u00eb t\u00eb till\u00eb zakonisht zgjidhet me grupe rregullash t\u00eb p\u00ebrcaktuara nga ekspert\u00ebt, t\u00eb cilat jan\u00eb shum\u00eb t\u00eb punshme n\u00eb zhvillim dhe mb\u00ebshtetje. Eficienca e k\u00ebsaj automatizimi, n\u00eb var\u00ebsi t\u00eb nivelit t\u00eb kompleksitetit t\u00eb detyr\u00ebs, mund t\u00eb arrij\u00eb 20\u201330%. Natyrisht, lind ideja q\u00eb \u00ebsht\u00eb m\u00eb e dobishme t\u00eb implementohet nj\u00eb moduli i inteligjenc\u00ebs artificiale \u2014 nj\u00eb model i zhvilluar me ndihm\u00ebn e m\u00ebsimit t\u00eb makineris\u00eb, i cili:<\/p>\n<ul>\n<li>\u00ebsht\u00eb n\u00eb gjendje t\u00eb procesoj\u00eb pa ndihm\u00ebn e operatorit nj\u00eb num\u00ebr m\u00eb t\u00eb madh k\u00ebrkesash (var\u00ebsisht nga tema, n\u00eb disa raste efikasiteti mund t\u00eb arrij\u00eb 70\u201380%);\n<\/li>\n<li>p\u00ebrtypihet m\u00eb mir\u00eb me formulime jo standarde n\u00eb bised\u00eb \u2014 di t\u00eb p\u00ebrcaktoj\u00eb q\u00ebllimin, d\u00ebshir\u00ebn reale t\u00eb p\u00ebrdoruesit p\u00ebrmes nj\u00eb k\u00ebrkese t\u00eb paqart\u00eb;\n<\/li>\n<li>di t\u00eb p\u00ebrcaktoj\u00eb kur p\u00ebrgjigjja e modelit \u00ebsht\u00eb adekuate, dhe kur si \u2018vet\u00ebdijshm\u00ebri\u2019 p\u00ebr k\u00ebt\u00eb p\u00ebrgjigje ka dyshime dhe nevojitet t\u00eb b\u00ebhet nj\u00eb pyetje sqaruese ose t\u00eb kalojm\u00eb te operatori;\n<\/li>\n<li>mund t\u00eb rihape automatikisht (n\u00eb vend t\u00eb nj\u00eb grupi zhvilluesish q\u00eb vazhdimisht adaptojn\u00eb dhe korrigjojn\u00eb skenar\u00ebt e p\u00ebrgjigjeve, modeli rihapet nga nj\u00eb specialist n\u00eb Shkenc\u00ebn e t\u00eb Dh\u00ebnave, duke p\u00ebrdorur bibliotekat p\u00ebrkat\u00ebse t\u00eb m\u00ebsimit t\u00eb makineris\u00eb).\u00a0\n<\/li>\n<\/ul>\n<p>\n<img decoding=\"async\" alt=\"MLOps: DevOps n\u00eb bot\u00ebn e Machine Learning\" src=\"\/wp-content\/uploads\/2020\/06\/9ef6356090a250df6cb212ab8bedb5f3.jpeg\" style=\"display:block;margin: 0 auto;\" \/><br \/>\n<br \/>\nSi ta b\u00ebjm\u00eb t\u00eb funksionoj\u00eb nj\u00eb model kaq t\u00eb avancuar?\u00a0<\/p>\n<p>Si\u00e7 ndodh me zgjidhjen e \u00e7do detyre tjet\u00ebr, p\u00ebrpara se t\u00eb zhvillohet nj\u00eb modul i till\u00eb, \u00ebsht\u00eb e nevojshme t\u00eb p\u00ebrcaktohet procesi i biznesit dhe t\u00eb p\u00ebrshkruhet formalisht detyra specifike q\u00eb do t\u00eb zgjidhim me ndihm\u00ebn e metod\u00ebs s\u00eb m\u00ebsimit t\u00eb makineris\u00eb. N\u00eb k\u00ebt\u00eb pik\u00eb fillon procesi i operacionalizimit, i sh\u00ebnuar me akronimin Ops.\u00a0<\/p>\n<p>Hapi tjet\u00ebr \u00ebsht\u00eb q\u00eb specialisti i Data science, n\u00eb bashk\u00ebpunim me inxhinierin e t\u00eb dh\u00ebnave, kontrollon disponueshm\u00ebrin\u00eb dhe mjaftueshm\u00ebrin\u00eb e t\u00eb dh\u00ebnave dhe hipotez\u00ebn e biznesit p\u00ebr efektshm\u00ebrin\u00eb e ideve t\u00eb biznesit, duke zhvilluar nj\u00eb prototip t\u00eb modelit dhe duke kontrolluar efektshm\u00ebrin\u00eb e tij t\u00eb v\u00ebrtet\u00eb. Vet\u00ebm pasi q\u00eb kjo t\u00eb konfirmohet nga biznesi, mund t\u00eb filloj\u00eb kalimi nga zhvillimi i modelit n\u00eb integrimin e tij n\u00eb sistemet q\u00eb realizojn\u00eb nj\u00eb proces t\u00eb caktuar biznesi. Planifikimi i plot\u00eb i integrimit, nj\u00eb kuptim i thell\u00eb n\u00eb \u00e7do hap se si do t\u00eb p\u00ebrdoret modeli dhe \u00e7far\u00eb efekti ekonomik do t\u00eb sjell\u00eb, \u00ebsht\u00eb nj\u00eb moment thelb\u00ebsor n\u00eb proceset e integrimit t\u00eb qasjeve MLOps n\u00eb peizazhin teknologjik t\u00eb kompanis\u00eb.<\/p>\n<p>Me zhvillimin e teknologjive t\u00eb AI-s\u00eb, numri dhe llojet e nj\u00eb gam\u00eb t\u00eb gjer\u00eb detyrash q\u00eb mund t\u00eb zgjidhen p\u00ebrmes m\u00ebsimit t\u00eb makinerive po rriten me shpejt\u00ebsi. \u00c7do proces biznesi i till\u00eb p\u00ebrfaq\u00ebson nj\u00eb kursim p\u00ebr kompanin\u00eb p\u00ebrmes automatizimit t\u00eb pun\u00ebs p\u00ebr pozita masive (qendrat telefonike, kontrolli dhe renditja e dokumenteve, etj.), zgjerimin e baz\u00ebs s\u00eb klient\u00ebve p\u00ebrmes shtimit t\u00eb funksioneve t\u00eb reja t\u00ebrheq\u00ebse dhe t\u00eb p\u00ebrshtatshme, kursimi i mjeteve p\u00ebrmes p\u00ebrdorimit optimal t\u00eb tyre dhe ri-ndarjes s\u00eb burimeve dhe shum\u00eb t\u00eb tjera. N\u00eb fund t\u00eb fundit, \u00e7do proces \u00ebsht\u00eb me orientim ndaj krijimit t\u00eb vler\u00ebs dhe, si rezultat, duhet t\u00eb sjell\u00eb nj\u00eb efekt ekonomik t\u00eb caktuar. K\u00ebtu \u00ebsht\u00eb shum\u00eb e r\u00ebnd\u00ebsishme t\u00eb formuloni sakt\u00eb iden\u00eb e biznesit dhe t\u00eb llogaritni fitimin e pritur nga implementimi i modelit n\u00eb struktur\u00ebn e krijimit t\u00eb vler\u00ebs s\u00eb kompanis\u00eb. Ka raste kur implementimi i modelit nuk justifikohet dhe koha e shpenzuar nga specialist\u00ebt e m\u00ebsimit t\u00eb makinerive \u00ebsht\u00eb m\u00eb e shtrenjt\u00eb se sa vendi i pun\u00ebs t\u00eb operatorit q\u00eb e realizon k\u00ebt\u00eb detyr\u00eb. Prandaj, \u00ebsht\u00eb thelb\u00ebsore t\u00eb p\u00ebrpiqemi t\u00eb identifikojm\u00eb t\u00eb tilla raste n\u00eb fazat e hershme t\u00eb krijimit t\u00eb sistemeve t\u00eb AI.<\/p>\n<p>Prandaj, fitimi nga modeli fillon t\u00eb mb\u00ebrrij\u00eb vet\u00ebm kur n\u00eb procesin e MLOps \u00ebsht\u00eb formuluar sakt\u00eb detyra biznesore, jan\u00eb vendosur prioritetet dhe n\u00eb fazat e hershme t\u00eb zhvillimit \u00ebsht\u00eb formuluar procesi i implementimit t\u00eb modelit n\u00eb sistem.<\/p>\n<h2>Procesi i ri \u2013 sfida t\u00eb reja<\/h2>\n<p>\nNj\u00eb p\u00ebrgjigje gjith\u00ebp\u00ebrfshir\u00ebse n\u00eb pyetjen themelore t\u00eb biznesit se sa t\u00eb aplikueshme jan\u00eb modelet ML p\u00ebr zgjidhjen e problemeve, si dhe \u00e7\u00ebshtja e p\u00ebrgjithshme e besimit n\u00eb AI - \u00ebsht\u00eb nj\u00eb nga sfidat kryesore n\u00eb procesin e zhvillimit dhe zbatimit t\u00eb qasjeve MLOps. Fillimisht, biznesi e percepton me skeptikiz\u00ebm zbatimin e m\u00ebsimit t\u00eb makinerive n\u00eb procese - \u00ebsht\u00eb e v\u00ebshtir\u00eb t\u00eb mb\u00ebshtetesh n\u00eb modele n\u00eb ato vende ku m\u00eb par\u00eb zakonisht punonin njer\u00ebzit. P\u00ebr biznesin, programet paraqiten si<\/p>\n<p>Sfid\u00eb e dyt\u00eb \u00ebsht\u00eb vler\u00ebsimi dhe llogaritja e rreziqeve t\u00eb modelit gjat\u00eb zbatimit t\u00eb modelit t\u00eb m\u00ebsimit t\u00eb makinerive. N\u00ebse edhe nj\u00eb njeri nuk mund t\u00eb p\u00ebrgjigjet me 100% siguri n\u00eb pyetjen n\u00ebse ajo fustan ishte e bardh\u00eb apo e kalt\u00ebr, at\u00ebher\u00eb edhe inteligjenca artificiale ka t\u00eb drejt\u00eb t\u00eb gaboj\u00eb. Gjithashtu, duhet marr\u00eb parasysh se p\u00ebr shembuj, t\u00eb dh\u00ebnat mund t\u00eb ndryshojn\u00eb me kalimin e koh\u00ebs dhe modelet duhet t\u00eb m\u00ebsohen m\u00eb tej p\u00ebr t\u00eb dh\u00ebn\u00eb rezultate mjaft t\u00eb sakta. P\u00ebr t\u00eb mos d\u00ebmtuar procesin e biznesit, \u00ebsht\u00eb e nevojshme t\u00eb menaxhohen rreziqet e modelit dhe t\u00eb monitorohet puna e modelit, duke e m\u00ebsuar at\u00eb rregullisht me t\u00eb dh\u00ebna t\u00eb reja.<\/p>\n<p><img decoding=\"async\" alt=\"MLOps: DevOps n\u00eb bot\u00ebn e Machine Learning\" src=\"\/wp-content\/uploads\/2020\/06\/cb97ff0eab95d81e17bf76aad3c3994e.jpeg\" style=\"display:block;margin: 0 auto;\" \/><br \/>\n<br \/>\nPor pas faz\u00ebs s\u00eb par\u00eb t\u00eb mosbesimit fillon t\u00eb shfaqet efekti i kund\u00ebrt. Sa m\u00eb shum\u00eb modele t\u00eb zbatohen me sukses n\u00eb procese, aq m\u00eb shum\u00eb rritet apetiti i biznesit p\u00ebr p\u00ebrdorimin e inteligjenc\u00ebs artificiale - gjenden vazhdimisht detyra t\u00eb reja q\u00eb mund t\u00eb zgjidhen me metodat e m\u00ebsimit t\u00eb makinerive. \u00c7do detyr\u00eb nis nj\u00eb proces t\u00eb t\u00ebr\u00eb, q\u00eb k\u00ebrkon kompetenca t\u00eb ndryshme:<\/p>\n<ul>\n<li>inxhinier\u00ebt e t\u00eb dh\u00ebnave p\u00ebrgatisin dhe p\u00ebrpunojn\u00eb t\u00eb dh\u00ebnat;\n<\/li>\n<li>shkenc\u00ebtar\u00ebt e t\u00eb dh\u00ebnave aplikojn\u00eb mjetet e m\u00ebsimit t\u00eb makinerive dhe zhvillojn\u00eb modelin;\n<\/li>\n<li>IT implementon modelin n\u00eb sistem;\n<\/li>\n<li>Inxhinieri ML p\u00ebrcakton se si ta integrojm\u00eb k\u00ebt\u00eb model n\u00eb procesin, cilat mjete IT t\u00eb p\u00ebrdorin n\u00eb var\u00ebsi t\u00eb k\u00ebrkesave p\u00ebr m\u00ebnyr\u00ebn e aplikimit t\u00eb modelit duke marr\u00eb parasysh fluksin e k\u00ebrkesave, koh\u00ebn e reagimit, etj.\u00a0\n<\/li>\n<li>Arkitekti ML projekton se si mund t\u00eb realizohet fizikisht produkti softuerik n\u00eb sistemin industrial.\n<\/li>\n<\/ul>\n<p>\nCikli i t\u00ebr\u00eb k\u00ebrkon nj\u00eb num\u00ebr t\u00eb madh specialist\u00ebsh t\u00eb kualifikuar. N\u00eb nj\u00eb pik\u00eb t\u00eb caktuar t\u00eb zhvillimit dhe t\u00eb dep\u00ebrtimit t\u00eb modeleve ML n\u00eb proceset e biznesit, ndodh q\u00eb t\u00eb rritet numri i specialist\u00ebve proporcionalisht me rritjen e numrit t\u00eb detyrave b\u00ebhet e shtrenjt\u00eb dhe e paefektshme. Prandaj, lind nevoja p\u00ebr automatizimin e procesit MLOps \u2014 p\u00ebrcaktimi i disa klasave standarde t\u00eb detyrave t\u00eb m\u00ebsimit t\u00eb makineris\u00eb, zhvillimi i pipeline-ve tipike t\u00eb p\u00ebrpunimit t\u00eb t\u00eb dh\u00ebnave dhe rinovimi i modeleve. N\u00eb nj\u00eb portret ideal p\u00ebr zgjidhjen e k\u00ebtyre detyrave k\u00ebrkohen profesionist\u00eb q\u00eb zot\u00ebrojn\u00eb nj\u00ebsoj mir\u00eb kompetencat n\u00eb kryq\u00ebzimin e BigData, Data Science, DevOps dhe IT. Prandaj, problemi m\u00eb i madh n\u00eb industrin\u00eb e Data Science dhe sfida m\u00eb e madhe n\u00eb organizimin e proceseve MLOps \u00ebsht\u00eb mungesa e k\u00ebsaj kompetence n\u00eb tregun aktual t\u00eb fuqis\u00eb pun\u00ebtore. Specialist\u00ebt q\u00eb plot\u00ebsojn\u00eb k\u00ebto k\u00ebrkesa aktualisht jan\u00eb t\u00eb rrall\u00eb n\u00eb tregun e pun\u00ebs dhe vler\u00ebsohen si arin.<\/p>\n<h2>P\u00ebrsa i p\u00ebrket kompetencave<\/h2>\n<p>\nTeoria thot\u00eb se t\u00eb gjitha detyrat e MLOps mund t\u00eb zgjidhen me mjetet klasike DevOps dhe pa u mb\u00ebshtetur n\u00eb zgjerimin e modelit t\u00eb rolit t\u00eb specializuar. Si\u00e7 e theksuam m\u00eb sip\u00ebr, data scientist nuk duhet t\u00eb jet\u00eb vet\u00ebm nj\u00eb matematicien dhe specialist i analiz\u00ebs s\u00eb t\u00eb dh\u00ebnave, por gjithashtu nj\u00eb guru i t\u00ebr\u00eb pipeline-it \u2014 ai merret me zhvillimin e arkitektur\u00ebs, programimin e modeleve n\u00eb disa gjuh\u00eb n\u00eb var\u00ebsi t\u00eb arkitektur\u00ebs, p\u00ebrgatitjen e vitrin\u00ebs s\u00eb t\u00eb dh\u00ebnave dhe shp\u00ebrndarjen e aplikacionit. Megjithat\u00eb, krijimi i infrastruktur\u00ebs teknologjike q\u00eb realizohet n\u00eb procesin e vazhduesh\u00ebm MLOps z\u00eb deri n\u00eb 80% t\u00eb koh\u00ebs s\u00eb pun\u00ebs, q\u00eb do t\u00eb thot\u00eb se nj\u00eb matematicien i kualifikuar, i cili \u00ebsht\u00eb nj\u00eb Data Scientist i mir\u00eb, do t'i kushtonte vet\u00ebm 20% t\u00eb koh\u00ebs profesionit t\u00eb tij. Prandaj, ndarja e rolit t\u00eb specialist\u00ebve q\u00eb kryejn\u00eb procesin e integrimit t\u00eb modeleve t\u00eb m\u00ebsimit t\u00eb makineris\u00eb b\u00ebhet jetike.\u00a0<\/p>\n<p>Sa sa e qart\u00eb se sa detajisht duhet t\u00eb ndahen rolet varet nga shkalla e nd\u00ebrmarrjes. \u00cbsht\u00eb nj\u00eb gj\u00eb kur n\u00eb nj\u00eb startup ka nj\u00eb specialist, pun\u00ebtor i energjis\u00eb, q\u00eb \u00ebsht\u00eb vet\u00eb inxhinier, arkitekt dhe DevOps. \u00cbsht\u00eb krejt\u00ebsisht di\u00e7ka tjet\u00ebr kur n\u00eb nj\u00eb kompani t\u00eb madhe t\u00eb gjitha proceset e zhvillimit t\u00eb modeleve jan\u00eb p\u00ebrqendruar te disa specialist\u00eb t\u00eb Data Science t\u00eb nivelit t\u00eb lart\u00eb, nd\u00ebrsa programi apo specialisti i baz\u00ebs s\u00eb t\u00eb dh\u00ebnave \u2013 \u00ebsht\u00eb nj\u00eb kompetenc\u00eb m\u00eb e zakonshme dhe m\u00eb pak e shtrenjt\u00eb n\u00eb tregun e pun\u00ebs \u2013 mund t\u00eb marr\u00eb p\u00ebrsip\u00ebr pjes\u00ebn m\u00eb t\u00eb madhe t\u00eb detyrave rutin\u00eb.<\/p>\n<p>Prandaj, nga vendi ku kalon kufiri n\u00eb zgjedhjen e specialist\u00ebve p\u00ebr t\u00eb siguruar procesin MLOps dhe si \u00ebsht\u00eb organizuar procesi i operacionalizimit t\u00eb modeleve t\u00eb zhvilluara, varet drejtp\u00ebrdrejt shpejt\u00ebsia dhe cil\u00ebsia e modeleve t\u00eb zhvilluara, produktiviteti i ekipit dhe mikroklima brenda tij.<\/p>\n<h2>\u00c7far\u00eb \u00ebsht\u00eb b\u00ebr\u00eb deri tani nga ekipi yn\u00eb<\/h2>\n<p>\nPak koh\u00eb m\u00eb par\u00eb filluam t\u00eb nd\u00ebrtojm\u00eb struktur\u00ebn e kompetencave dhe proceset MLOps. Por tashm\u00eb \u00ebsht\u00eb n\u00eb faz\u00ebn e testimit MVP projektet tona p\u00ebr menaxhimin e ciklit t\u00eb jet\u00ebs s\u00eb modeleve dhe p\u00ebr p\u00ebrdorimin e modeleve si sh\u00ebrbim.<\/p>\n<p>Po ashtu, ne kemi p\u00ebrcaktuar struktur\u00ebn optimale p\u00ebr nj\u00eb kompani t\u00eb madhe dhe struktur\u00ebn organizative t\u00eb bashk\u00ebveprimit midis t\u00eb gjith\u00eb pjes\u00ebmarr\u00ebsve t\u00eb procesit. Jan\u00eb organizuar ekipet Agile, q\u00eb zgjidhin detyra p\u00ebr t\u00eb gjith\u00eb spektrin e bizneseve, si dhe \u00ebsht\u00eb p\u00ebrmir\u00ebsuar procesi i bashk\u00ebveprimit me ekipet projektuale p\u00ebr krijimin e platformave, infrastruktur\u00ebs q\u00eb \u00ebsht\u00eb themeli i nd\u00ebrtes\u00ebs MLOps q\u00eb po nd\u00ebrtohet.<\/p>\n<h2>Pyetjet p\u00ebr t\u00eb ardhmen<\/h2>\n<p>\nMLOps \u00ebsht\u00eb nj\u00eb drejtim n\u00eb zhvillim, i cili p\u00ebrjeton munges\u00eb kompetencash dhe n\u00eb t\u00eb ardhmen do t\u00eb marr\u00eb p\u00ebrmasa. Deri at\u00ebher\u00eb, \u00ebsht\u00eb m\u00eb e mira t\u00eb mb\u00ebshtetemi n\u00eb p\u00ebrvojat dhe praktikat DevOps. Q\u00ebllimi kryesor i MLOps \u00ebsht\u00eb p\u00ebrdorimi m\u00eb efektiv i modeleve ML p\u00ebr zgjidhjen e problemeve t\u00eb biznesit. Por me k\u00ebt\u00eb lindin shum\u00eb pyetje:<\/p>\n<ul>\n<li>Si t\u00eb reduktojm\u00eb koh\u00ebn p\u00ebr t\u00eb lan\u00e7uar modelet n\u00eb prodhim?\n<\/li>\n<li>Si t\u00eb ulin frictionet burokratike midis ekipeve me kompetenca t\u00eb ndryshme dhe t\u00eb rrisin fokusin n\u00eb bashk\u00ebpunim?\n<\/li>\n<li>Si t\u00eb monitorojm\u00eb modelet, t\u00eb menaxhojm\u00eb versionet dhe t\u00eb organizojm\u00eb nj\u00eb monitorim efektiv?\n<\/li>\n<li>Si t\u00eb krijojm\u00eb v\u00ebrtet nj\u00eb cik\u00ebl t\u00eb ciklik t\u00eb jet\u00ebs p\u00ebr nj\u00eb model t\u00eb modern ML?\n<\/li>\n<li>Si t\u00eb standardizojm\u00eb procesin e m\u00ebsimit t\u00eb makin\u00ebs?\n<\/li>\n<\/ul>\n<p>\nP\u00ebrgjigjet e k\u00ebtyre pyetjeve do t\u00eb p\u00ebrcaktojn\u00eb shum\u00eb nga sa shpejt MLOps do t\u00eb zbuloj\u00eb potencialin e saj plot\u00ebsisht.<br \/>\n<br \/>Burimi: <a content=\"nofollow\" rel=\"nofollow\" href=\"https:\/\/habr.com\/ru\/company\/vtb\/blog\/508012\/\">habr.com<\/a> <\/p>","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"excerpt":{"rendered":"<p>\u0412 2018 \u0433\u043e\u0434\u0443 \u0432 \u043f\u0440\u043e\u0444\u0435\u0441\u0441\u0438\u043e\u043d\u0430\u043b\u044c\u043d\u044b\u0445 \u043a\u0440\u0443\u0433\u0430\u0445 \u0438 \u043d\u0430 \u0442\u0435\u043c\u0430\u0442\u0438\u0447\u0435\u0441\u043a\u0438\u0445 \u043a\u043e\u043d\u0444\u0435\u0440\u0435\u043d\u0446\u0438\u044f\u0445, \u043f\u043e\u0441\u0432\u044f\u0449\u0435\u043d\u043d\u044b\u0445 AI, \u043f\u043e\u044f\u0432\u0438\u043b\u043e\u0441\u044c \u043f\u043e\u043d\u044f\u0442\u0438\u0435 MLOps, \u043a\u043e\u0442\u043e\u0440\u043e\u0435 \u0431\u044b\u0441\u0442\u0440\u043e \u0437\u0430\u043a\u0440\u0435\u043f\u0438\u043b\u043e\u0441\u044c \u0432 \u043e\u0442\u0440\u0430\u0441\u043b\u0438 \u0438 \u0441\u0435\u0439\u0447\u0430\u0441 \u0440\u0430\u0437\u0432\u0438\u0432\u0430\u0435\u0442\u0441\u044f \u043a\u0430\u043a \u0441\u0430\u043c\u043e\u0441\u0442\u043e\u044f\u0442\u0435\u043b\u044c\u043d\u043e\u0435 \u043d\u0430\u043f\u0440\u0430\u0432\u043b\u0435\u043d\u0438\u0435. \u0412 \u043f\u0435\u0440\u0441\u043f\u0435\u043a\u0442\u0438\u0432\u0435 MLOps \u043c\u043e\u0436\u0435\u0442 \u0441\u0442\u0430\u0442\u044c \u043e\u0434\u043d\u043e\u0439 \u0438\u0437 \u043d\u0430\u0438\u0431\u043e\u043b\u0435\u0435 \u0432\u043e\u0441\u0442\u0440\u0435\u0431\u043e\u0432\u0430\u043d\u043d\u044b\u0445 \u0441\u0444\u0435\u0440 \u0432 IT. \u0427\u0442\u043e \u0436\u0435 \u044d\u0442\u043e \u0442\u0430\u043a\u043e\u0435 \u0438 \u0441 \u0447\u0435\u043c \u0435\u0433\u043e \u0435\u0434\u044f\u0442, \u0440\u0430\u0437\u0431\u0438\u0440\u0430\u0435\u043c\u0441\u044f \u043f\u043e\u0434 \u043a\u0430\u0442\u043e\u043c. \u0427\u0442\u043e \u0442\u0430\u043a\u043e\u0435 MLOps MLOps (\u043e\u0431\u044a\u0435\u0434\u0438\u043d\u0435\u043d\u0438\u0435 [&hellip;]<\/p>\n","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"author":1,"featured_media":86524,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[688],"tags":[],"class_list":["post-86523","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-administrirovanie"],"aioseo_notices":[],"aioseo_head":"\n\t\t<!-- All in One SEO 5.0.2.1 - aioseo.com -->\n\t<meta name=\"description\" content=\"\u0412 2018 \u0433\u043e\u0434\u0443 \u0432 \u043f\u0440\u043e\u0444\u0435\u0441\u0441\u0438\u043e\u043d\u0430\u043b\u044c\u043d\u044b\u0445 \u043a\u0440\u0443\u0433\u0430\u0445 \u0438 \u043d\u0430 \u0442\u0435\u043c\u0430\u0442\u0438\u0447\u0435\u0441\u043a\u0438\u0445 \u043a\u043e\u043d\u0444\u0435\u0440\u0435\u043d\u0446\u0438\u044f\u0445, \u043f\u043e\u0441\u0432\u044f\u0449\u0435\u043d\u043d\u044b\u0445 AI, \u043f\u043e\u044f\u0432\u0438\u043b\u043e\u0441\u044c \u043f\u043e\u043d\u044f\u0442\u0438\u0435 MLOps, \u043a\u043e\u0442\u043e\u0440\u043e\u0435 \u0431\u044b\u0441\u0442\u0440\u043e \u0437\u0430\u043a\u0440\u0435\u043f\u0438\u043b\u043e\u0441\u044c \u0432 \u043e\u0442\u0440\u0430\u0441\u043b\u0438 \u0438 \u0441\u0435\u0439\u0447\u0430\u0441 \u0440\u0430\u0437\u0432\u0438\u0432\u0430\u0435\u0442\u0441\u044f \u043a\u0430\u043a \u0441\u0430\u043c\u043e\u0441\u0442\u043e\u044f\u0442\u0435\u043b\u044c\u043d\u043e\u0435 \u043d\u0430\u043f\u0440\u0430\u0432\u043b\u0435\u043d\u0438\u0435.\" \/>\n\t<meta name=\"robots\" content=\"max-image-preview:large\" \/>\n\t<meta name=\"author\" content=\"Yuri Gagarin\"\/>\n\t<link rel=\"canonical\" href=\"https:\/\/prohoster.info\/sq\/blog\/administrirovanie\/mlops-devops-v-mire-machine-learning\" \/>\n\t<meta name=\"generator\" content=\"All in One SEO (AIOSEO) 5.0.2.1\" \/>\n\t\t<meta property=\"og:locale\" content=\"sq_AL\" \/>\n\t\t<meta property=\"og:site_name\" content=\"ProHoster | \u041a\u0443\u043f\u0438\u0442\u044c \u043d\u0430\u0434\u0435\u0436\u043d\u044b\u0439 \u0445\u043e\u0441\u0442\u0438\u043d\u0433 \u0434\u043b\u044f \u0441\u0430\u0439\u0442\u043e\u0432 \u0441 \u0437\u0430\u0449\u0438\u0442\u043e\u0439 \u043e\u0442 DDoS, VPS VDS \u0441\u0435\u0440\u0432\u0435\u0440\u044b\" \/>\n\t\t<meta property=\"og:type\" content=\"article\" \/>\n\t\t<meta property=\"og:title\" content=\"\ud83e\udd47MLOps: DevOps \u0432 \u043c\u0438\u0440\u0435 Machine Learning | ProHoster\" \/>\n\t\t<meta property=\"og:description\" content=\"\u0412 2018 \u0433\u043e\u0434\u0443 \u0432 \u043f\u0440\u043e\u0444\u0435\u0441\u0441\u0438\u043e\u043d\u0430\u043b\u044c\u043d\u044b\u0445 \u043a\u0440\u0443\u0433\u0430\u0445 \u0438 \u043d\u0430 \u0442\u0435\u043c\u0430\u0442\u0438\u0447\u0435\u0441\u043a\u0438\u0445 \u043a\u043e\u043d\u0444\u0435\u0440\u0435\u043d\u0446\u0438\u044f\u0445, \u043f\u043e\u0441\u0432\u044f\u0449\u0435\u043d\u043d\u044b\u0445 AI, \u043f\u043e\u044f\u0432\u0438\u043b\u043e\u0441\u044c \u043f\u043e\u043d\u044f\u0442\u0438\u0435 MLOps, \u043a\u043e\u0442\u043e\u0440\u043e\u0435 \u0431\u044b\u0441\u0442\u0440\u043e \u0437\u0430\u043a\u0440\u0435\u043f\u0438\u043b\u043e\u0441\u044c \u0432 \u043e\u0442\u0440\u0430\u0441\u043b\u0438 \u0438 \u0441\u0435\u0439\u0447\u0430\u0441 \u0440\u0430\u0437\u0432\u0438\u0432\u0430\u0435\u0442\u0441\u044f \u043a\u0430\u043a \u0441\u0430\u043c\u043e\u0441\u0442\u043e\u044f\u0442\u0435\u043b\u044c\u043d\u043e\u0435 \u043d\u0430\u043f\u0440\u0430\u0432\u043b\u0435\u043d\u0438\u0435.\" \/>\n\t\t<meta property=\"og:url\" content=\"https:\/\/prohoster.info\/sq\/blog\/administrirovanie\/mlops-devops-v-mire-machine-learning\" \/>\n\t\t<meta property=\"og:image\" content=\"https:\/\/prohoster.info\/wp-content\/uploads\/2021\/11\/logo-350.jpg\" \/>\n\t\t<meta property=\"og:image:secure_url\" content=\"https:\/\/prohoster.info\/wp-content\/uploads\/2021\/11\/logo-350.jpg\" \/>\n\t\t<meta property=\"og:image:width\" content=\"350\" \/>\n\t\t<meta property=\"og:image:height\" content=\"350\" \/>\n\t\t<meta property=\"article:published_time\" content=\"2020-06-26T05:42:14+00:00\" \/>\n\t\t<meta property=\"article:modified_time\" content=\"2020-06-26T05:42:14+00:00\" \/>\n\t\t<meta property=\"article:publisher\" content=\"https:\/\/www.facebook.com\/prohoster\" \/>\n\t\t<meta property=\"article:author\" content=\"https:\/\/www.facebook.com\/prohoster\" \/>\n\t\t<!-- All in One SEO -->\n\n","aioseo_head_json":{"title":"\ud83e\udd47MLOps: DevOps n\u00eb bot\u00ebn e Machine Learning | ProHoster","description":"N\u00eb vitin 2018, n\u00eb rrethin profesional dhe n\u00eb konferencat tematike t\u00eb dedikuara p\u00ebr AI, u shpall koncepti MLOps, i cili shpejt u konsolidua n\u00eb industri dhe tani po zhvillohet si nj\u00eb drejtim i pavarur.","canonical_url":"https:\/\/prohoster.info\/sq\/blog\/administrirovanie\/mlops-devops-v-mire-machine-learning","robots":"max-image-preview:large","keywords":"","webmasterTools":{"miscellaneous":""},"schema":null,"og:locale":"sq_AL","og:site_name":"ProHoster | \u041a\u0443\u043f\u0438\u0442\u044c \u043d\u0430\u0434\u0435\u0436\u043d\u044b\u0439 \u0445\u043e\u0441\u0442\u0438\u043d\u0433 \u0434\u043b\u044f \u0441\u0430\u0439\u0442\u043e\u0432 \u0441 \u0437\u0430\u0449\u0438\u0442\u043e\u0439 \u043e\u0442 DDoS, VPS VDS \u0441\u0435\u0440\u0432\u0435\u0440\u044b","og:type":"article","og:title":"\ud83e\udd47MLOps: DevOps \u0432 \u043c\u0438\u0440\u0435 Machine Learning | ProHoster","og:description":"\u0412 2018 \u0433\u043e\u0434\u0443 \u0432 \u043f\u0440\u043e\u0444\u0435\u0441\u0441\u0438\u043e\u043d\u0430\u043b\u044c\u043d\u044b\u0445 \u043a\u0440\u0443\u0433\u0430\u0445 \u0438 \u043d\u0430 \u0442\u0435\u043c\u0430\u0442\u0438\u0447\u0435\u0441\u043a\u0438\u0445 \u043a\u043e\u043d\u0444\u0435\u0440\u0435\u043d\u0446\u0438\u044f\u0445, \u043f\u043e\u0441\u0432\u044f\u0449\u0435\u043d\u043d\u044b\u0445 AI, \u043f\u043e\u044f\u0432\u0438\u043b\u043e\u0441\u044c \u043f\u043e\u043d\u044f\u0442\u0438\u0435 MLOps, \u043a\u043e\u0442\u043e\u0440\u043e\u0435 \u0431\u044b\u0441\u0442\u0440\u043e \u0437\u0430\u043a\u0440\u0435\u043f\u0438\u043b\u043e\u0441\u044c \u0432 \u043e\u0442\u0440\u0430\u0441\u043b\u0438 \u0438 \u0441\u0435\u0439\u0447\u0430\u0441 \u0440\u0430\u0437\u0432\u0438\u0432\u0430\u0435\u0442\u0441\u044f \u043a\u0430\u043a \u0441\u0430\u043c\u043e\u0441\u0442\u043e\u044f\u0442\u0435\u043b\u044c\u043d\u043e\u0435 \u043d\u0430\u043f\u0440\u0430\u0432\u043b\u0435\u043d\u0438\u0435.","og:url":"https:\/\/prohoster.info\/sq\/blog\/administrirovanie\/mlops-devops-v-mire-machine-learning","og:image":"https:\/\/prohoster.info\/wp-content\/uploads\/2021\/11\/logo-350.jpg","og:image:secure_url":"https:\/\/prohoster.info\/wp-content\/uploads\/2021\/11\/logo-350.jpg","og:image:width":350,"og:image:height":350,"article:published_time":"2020-06-26T05:42:14+00:00","article:modified_time":"2020-06-26T05:42:14+00:00","article:publisher":"https:\/\/www.facebook.com\/prohoster","article:author":"https:\/\/www.facebook.com\/prohoster"},"aioseo_meta_data":{"post_id":"86523","title":null,"description":null,"keywords":null,"keyphrases":null,"primary_term":null,"canonical_url":null,"og_title":null,"og_description":null,"og_object_type":"default","og_image_type":"default","og_image_url":null,"og_image_width":null,"og_image_height":null,"og_image_custom_url":null,"og_image_custom_fields":null,"og_video":null,"og_custom_url":null,"og_article_section":null,"og_article_tags":null,"twitter_use_og":false,"twitter_card":"default","twitter_image_type":"default","twitter_image_url":null,"twitter_image_custom_url":null,"twitter_image_custom_fields":null,"twitter_title":null,"twitter_description":null,"schema":{"blockGraphs":[],"customGraphs":[],"default":{"data":{"Article":[],"Course":[],"Dataset":[],"FAQPage":[],"Movie":[],"Person":[],"Product":[],"ProductReview":[],"Car":[],"Recipe":[],"Service":[],"SoftwareApplication":[],"WebPage":[]},"graphName":"","isEnabled":true},"graphs":[]},"schema_type":null,"schema_type_options":null,"pillar_content":false,"robots_default":true,"robots_noindex":false,"robots_noarchive":false,"robots_nosnippet":false,"robots_nofollow":false,"robots_noimageindex":false,"robots_noodp":false,"robots_notranslate":false,"robots_max_snippet":null,"robots_max_videopreview":null,"robots_max_imagepreview":"large","priority":null,"frequency":null,"local_seo":null,"seo_analyzer_scan_date":null,"breadcrumb_settings":null,"limit_modified_date":false,"reviewed_by":null,"ai":null,"created":"2021-02-28 14:13:23","updated":"2022-09-28 09:22:54","focus_keyword":null,"additional_keywords":null,"truseo_locale":null},"gt_translate_keys":[{"key":"link","format":"url"}],"_links":{"self":[{"href":"https:\/\/prohoster.info\/sq\/wp-json\/wp\/v2\/posts\/86523","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/prohoster.info\/sq\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/prohoster.info\/sq\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/prohoster.info\/sq\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/prohoster.info\/sq\/wp-json\/wp\/v2\/comments?post=86523"}],"version-history":[{"count":0,"href":"https:\/\/prohoster.info\/sq\/wp-json\/wp\/v2\/posts\/86523\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/prohoster.info\/sq\/wp-json\/wp\/v2\/media\/86524"}],"wp:attachment":[{"href":"https:\/\/prohoster.info\/sq\/wp-json\/wp\/v2\/media?parent=86523"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/prohoster.info\/sq\/wp-json\/wp\/v2\/categories?post=86523"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/prohoster.info\/sq\/wp-json\/wp\/v2\/tags?post=86523"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}