Ama-DevOps ane-AI kanye ne-LLMOps: kusukela kuphayiphi kuya kumodeli ekhulumayo

Informatec Digital » Izinsiza » Ama-DevOps ane-AI kanye ne-LLMOps: kusukela kuphayiphi kuya kumodeli ekhulumayo

Ukufika kwe-AI ekhiqizayo kanye namamodeli ezilimi ezinkulu kushintshe ngokuphelele indlela isofthiwe eyakhiwe, esetshenziswa futhi esetshenziswa ngayo. Ukuba nezinto ezinhle akusanele. Amapayipi e-DevOps noma ngokusebenzisa ama-MLOps akudalaUma wethula i-LLM kulesi sibalo, ungena endaweni lapho imodeli ikhuluma khona, icabanga, iqamba izinto ezintsha, futhi ngezinye izikhathi iziphatha ngezindlela ezingalindelekile.

Kulesi simo esisha, Amaqembu adinga ukuhlanganisa i-DevOps, i-AI, kanye ne-LLMOps ukuze alawule yonke inkathi yokuphila kwezinhlelo zokusebenza ezisekelwe ku-LLM.Kusukela ekuhlolweni nasekuthuthukisweni okusheshayo kuya ekufakweni, ukuqapha, ukuphepha, kanye nokwenza ngcono izindleko, lesi sihloko siletha wonke lowo msindo emhlabeni futhi sikuqondise, isinyathelo ngesinyathelo, ukuthi ungawafaka kanjani ama-ChatOps, ama-DevOps, ama-MLOps, ama-GenAIOps, kanye nama-LLMOps emsebenzini wesimanje.

Kusukela ku-DevOps kanye ne-MLOps kuya ku-LLMOps: kungani imodeli ingasasebenzi

Iminyaka eminingi, okubaluleke kakhulu kumaqembu obunjiniyela bekuyi ukulethwa kwesofthiwe ngokuzenzakalelayo futhi kuncishiswe ukungezwani phakathi kwentuthuko kanye nengqalasizindaNgakho-ke i-DevOps yazalwa: ukuhlanganiswa okuqhubekayo, ukuthunyelwa okuqhubekayo, ingqalasizinda njengekhodi, ukubonwa, kanye nesiko lokubambisana elaqeda ukudluliselwa okungapheli phakathi kweminyango.

Lapho idatha iba yingxenye yomkhiqizo, yavela Ama-MLOp njengempendulo yesidingo sokuphinda kukhiqizwe futhi kulandelelwe amamodeli okufunda komshiniImikhuba efana nokuguqulwa kwedatha, ukuhlelwa kwepayipi lokuqeqesha, ukutholwa kokukhukhuleka, kanye nokuhlolwa okuqhubekayo kwamamodeli okubikezela kwabekwa ezingeni elifanayo.

Inkinga ukuthi Ama-LLM aphula okuningi kokucabanga okucashile ku-DevOps kanye nama-MLOpsAkuzona ama-API angaguquki noma imisebenzi elula ebuyisela inombolo eqondile: iphendula ngolimi lwemvelo, ixuba umongo, imiyalelo, amathuluzi kanye nedatha ngesikhathi sangempela, futhi ingakhiqiza imiphumela emibili ehlukene yokufaka okufanayo.

Lokhu kusho ukuthi Akwanele ukushintsha nje imodeli kanye nezisindo zayoKuyadingeka futhi ukulawula izikhuthazo, amathempulethi, izinqubomgomo zokuphepha ezichazayo, imikhawulo, amathuluzi axhunyiwe, umongo ofakiwe, ngisho nemithetho yebhizinisi elawula ukuziphatha kwesistimu.

Iyini i-LLMOps futhi iyixazulula kanjani ngempela?

Singabona ama-LLMOp njenge uhlaka lokusebenza oluvumela ukuthunyelwa, ukugcinwa, kanye nokulinganiswa kwezinhlelo zokusebenza ezisekelwe ku-LLM okuphephile, okulawulwayo, nokuqhubekayoKuyisambulela lapho i-DevOps isebenza khona, ama-MLOp, kanye namakhono amasha akhethekile kumamodeli okukhiqiza ahlala khona ndawonye.

Empeleni, Ama-LLMOp awagxili kakhulu “ekuqeqesheni imodeli ephelele” kodwa agxila kakhulu ekulawuleni ukuziphatha kwawo ekukhiqizeni.Kuhlanganisa indlela ukugeleza okusheshayo okuklanywe ngayo futhi okuhunyushwe ngayo, indlela ama-LLM axhunyaniswa ngayo nemithombo yedatha yangaphakathi, indlela izindleko zamathokheni kanye nokubambezeleka okubhekwa ngayo, kanye nendlela ubungozi obuphathelene nencazelo obuphathwa ngayo (ukungaqondi kahle, ukuvuza kolwazi, ukucwasa, izimpendulo ezinobuthi, njll.).

Izidingo ezisingathwa yi-LLMOps, kanye nezingahlanganyelwa yi-DevOps/MLOps zodwa, zifaka phakathi izici ezahlukahlukene njengokulandelelwa kwengxoxo, ukuhlolwa okuzenzakalelayo kwekhwalithi yempendulo, noma ukuqhathaniswa kwe-A/B kwezinhlobo zokuziphathaAsikhulumi nje ngokunemba okuvamile, kodwa futhi nokungaguquguquki, ukuvumelana nebhizinisi, kanye nokuphepha.

Futhi, Izindleko azisekho kuphela ekuqeqeshweni nasekusingatheni imodeliIsaziso ngasinye, umongo ngamunye owandisiwe, kanye nocingo ngalunye oluhambisanayo kudala ukusetshenziswa kwe-GPU noma kwethokheni kuma-API ezentengiselwano. Ngaphandle kwesendlalelo se-LLMOps sokwenza lokhu kusetshenziswa kubonakale futhi sikuxhume kumishini, izinsizakalo, kanye nezimo zokusetshenziswa, umthethosivivinywa ukhula ngendlela engalindelekile.

Ama-ChatOps + ama-LLMOps + ama-DevOps: imisebenzi iba yingxoxo

Enye yezindlela ezinamandla kakhulu ukuhlanganiswa kwe-ChatOps kanye ne-LLMOps ngaphakathi kwesiko le-DevOpsEsikhundleni sokukhawulelwa kumadeshibhodi, izikripthi, kanye namapayipi, amaqembu aqala ukusebenzisa ingxenye enkulu yesistimu kusuka eziteshini zengxoxo ezifana ne-Slack, i-Microsoft Teams, noma i-Discord.

I-ChatOps iphakamisa ukuthi imisebenzi yansuku zonke (ukuthunyelwa, imibuzo yelogi, ukuqalisa kabusha, izinguquko zokucushwa) kwenziwa yi-bots ngaphakathi kwesiteshi sokuxhumana uqobo, ngokusobala eqenjini lonke. Yonke imiyalo, isenzo, kanye nomphumela kuqoshwa engxoxweni.

Uma i-LLM ingezwa kuleyo ndlela, kuvela ungqimba olusha lobuhlakani: Ama-Chatbot aqonda ulimi lwemvelo, ahumushe izinhloso, futhi angenza imiyalo eyinkimbinkimbi noma ahlaziye izimo ngaphandle kokuthi umqhubi adinge ukukhumbula yonke imibhalo noma ifulegi eliqondile.

Izibonelo ezijwayelekile zalokhu kuhlangana zifaka phakathi lokho i-bot, enikwe amandla yi-LLM, ifunda izibalo ze-Prometheus kanye namalogi e-Loki Uma umuntu ebhala ukuthi "inkonzo yeqembu X ihamba kancane" futhi iphakamisa izenzo ezinjengokukhulisa amakhophi, ukwenza i-rollback, noma ukuqalisa izivivinyo ezithile, konke kuchazwe ngolimi lwemvelo.

Ezingeni lamasiko nokusebenza, lokhu kuhunyushwa kube Izinqumo ezisheshayo, ukungenelela okuncane ngesandla emisebenzini ephindaphindwayo, kanye nokuhlangenwe nakho okubushelelezi kwamaqembu e-DevOps, abaqala ekucimeni imililo njalo baye ekusebenzeni ekuthuthukisweni kwamasu.

Izimiso ezibalulekile zomjikelezo wokuphila we-LLM ekukhiqizeni

Ukuqhuba i-LLM engathi sína akuyona iphrojekthi eyenzeka kanye kuphela, umjikelezo oziphindaphindayo futhi lapho ushintsho ngalunye lungashintsha khona ukuziphatha kwesistimuNakuba inhlangano ngayinye iyivumelanisa nesimo sayo, ngokuvamile kunezigaba eziyisithupha ezinkulu eziphinde zihlangane.

Esokuqala isigaba sokuqeqeshwa noma sokuzivumelanisa nesimo semodeliLokhu kungasukela ekusebenziseni imodeli eyisisekelo kanye nokusebenzisa ukulungisa kahle, i-LoRa, noma amanye amasu okulungisa ngedatha yakho. Into ebalulekile lapha akukhona nje ukusebenza, kodwa ukushiya irekhodi eliphelele: amasethi edatha, izihlungi ezisetshenzisiwe, ama-hyperparameter, izinguqulo ze-tokenizer, izakhiwo ezivivinyiwe, njll.

Uma lesi sigaba singesokwenziwa futhi singabhalwanga phansi, imodeli izalwa ngaphandle kokubusaNgemva kwalokho, cishe ngeke kwenzeke ukuchaza ukuthi kungani isabela ngendlela esabela ngayo noma ukuphinda umphumela othize uma kudingeka ekuhlolweni.

Isigaba sesibili ukuthunyelwa, lapho imodeli iphuma khona elebhu bese iqala ukukhiqizwa. Kwa-LLMOps, lokhu akukhona nje "ukuyifaka esitsheni": Kufanele sinqume iyiphi ihadiwe okufanele isetshenzisweIndlela yokuphatha inkumbulo yezimo ezisebenza isikhathi eside, ukuthi iyiphi i-topology yeqoqo okufanele isetshenziswe, nokuthi ungayikhulisa kanjani ngokusekelwe kuthrafikhi ngaphandle kokubambezeleka okukhulu noma izindleko zingabi namandla okuzikhokhela.

Yilapho izinto ziqala khona ukusebenza ukuqapha okuqhubekayo okugxile ekuziphatheniAkwanele ukubheka i-CPU ne-RAM; kuyadingeka ukuqapha ikhwalithi yesimantiki yezimpendulo, ukuzinza kwesitayela, izinga lamaphutha, ukuvela kwezindleko ngethokheni ngayinye, ukubonakala kwezimpendulo eziyingozi noma ezingahambisani kanye nezinguquko ezikhathini zokuphendula ngaphansi kwamaphethini okusetshenziswa ahlukene.

Ezigabeni zakamuva, imisebenzi yokwenza ngcono kanye nokulungisa kahle iyaqhutshwa: thinta izixwayiso, lungisa i-RAG, hlola izinhlobo zemodeli, linganisa, yenza ukuhlolwa kwe-A/B, shintsha izinqubomgomo zokuphepha eziyisisekelo, noma uthuthukise imithetho yebhizinisiKuyinqubo ecishe ibe yobuciko, lapho idatha, ubunjiniyela, kanye nebhizinisi zihlala phansi ndawonye ukuze zinqume ukuthi yini okufanele ibekwe phambili.

Ekugcineni, konke lokhu kuwela ngaphakathi izendlalelo zokuphepha kanye nokuphatha (ukulawula ukufinyelela, ukuhlola, ukuvimbela ukuvuza, imikhawulo yokusetshenziswa, ukuthobela imithetho) kanye nokubuyekezwa okuqhubekayo, lapho imodeli kanye nesistimu yayo yemvelo kulungiswa khona ukuze kuhambisane nezinguquko kudatha, imithethonqubo kanye nezidingo zangaphakathi.

I-GenAIOps kanye nendlela yokugeleza kwesaziso e-Azure

Ngaphakathi kwendawo yonke ye-LLMOps, kuneziphakamiso eziqondile kakhulu zokuhlela lo mjikelezo wokuphila. Enye yezinto ezithuthuke kakhulu endaweni yebhizinisi yi I-GenAIOps enokugeleza okusheshayo ku-Azure Machine Learning ehlanganiswe ne-Azure DevOps, okuphakamisa indlela ehlelekile kakhulu yokwakha izinhlelo zokusebenza ezisekelwe ku-LLM.

Ukugeleza okusheshayo akuyona nje umhleli we-prompt; ipulatifomu ephelele yokuklama, ukuhlola, ukuguqulela, kanye nokusebenzisa ukugeleza kokuxhumana kwe-LLM, kusukela ezimweni ezilula (isicelo esisodwa) kuya ekuhleleni okuyinkimbinkimbi okunama-node amaningi, amathuluzi angaphandle, izilawuli kanye nokuhlola okuzenzakalelayo.

Isici esibalulekile yilesi indawo yokugcina yokugeleza ephakathiesebenza njengomtapo wolwazi wenkampani. Esikhundleni sokuba iqembu ngalinye libe nezicelo zalo kumadokhumenti ahlukene noma ezindaweni zalo zokugcina, zihlanganiswa zibe yindawo yokugcina eyodwa ephethwe, enamagatsha acacile, izibuyekezo, kanye nomlando.

Ngaphezu kwalokho, ipulatifomu ingeza amakhono okuhlola ahlukahlukene kanye ne-hyperparameter: Kungenzeka ukuhlola inhlanganisela ehlukene yemiyalelo, amamodeli, izilungiselelo zokushisa, noma izinqubomgomo zokuphepha kuma-node amaningi okugeleza bese uqhathanisa imiphumela nezilinganiso ezicacile.

Ngokuphathelene nokuthunyelwa, i-GenAIOps ngokugeleza kwesaziso Ikhiqiza izithombe ze-Docker ezihlanganisa kokubili ukuhamba komsebenzi kanye neseshini yenqubo.Lokhu kulungele ukusebenza ezindaweni ezifana ne-Azure App Services, i-Kubernetes, noma izinqubo eziphethwe. Kusukela kulesi sisekelo, ukuthunyelwa kwe-A/B kuvunyelwe ukuqhathanisa izinguqulo zokugeleza ezindaweni zangempela.

Elinye ikhono lokuphatha ubudlelwano phakathi kwamasethi edatha kanye nokugeleza kwawo. Ukugeleza ngakunye kokuhlola kungasebenza ngamasethi amaningi edatha ajwayelekile kanye nokuhlolaLokhu kuvumela ukuqinisekisa ukuziphatha ezimweni ezahlukene ngaphambi kokubeka okuthile ezandleni zabasebenzisi bokugcina.

Ipulatifomu futhi ibhalisa ngokuzenzakalelayo izinguqulo ezintsha zamasethi edatha kanye nokugeleza kuphela uma kunezinguquko zangempela, futhi Ikhiqiza imibiko ephelele ngamafomethi afana ne-CSV ne-HTML. ukusekela izinqumo ezisekelwe kudatha, hhayi ekuqondeni.

Izigaba ezine ze-GenAIOps ngokugeleza kwesaziso

Indlela ye-GenAIOps ihlukanisa umjikelezo wokuphila ube yizigaba ezine ezihlukaniswe ngokucacile, okusiza ukugwema isiphithiphithi esivamile sokuthi "sizame izinto nge-AI bese sibona ukuthi kwenzekani".

Isigaba sokuqala, ukuqaliswa, sigxile ku Chaza inhloso yebhizinisi ngokunembile bese uqoqa izibonelo zedatha emeleLapha isakhiwo esiyisisekelo sokugeleza okusheshayo sichazwe futhi ukwakheka kwaklanywa, okuzobe sekucwengisiswa.

Esigabeni sokuhlola, ukugeleza kusetshenziswa kuleyo datha yesampula futhi Izinhlobo ezahlukene zezicelo, amamodeli, kanye nokucushwa kuyahlolwa.Le nqubo iphindaphindwa njalo kuze kube yilapho kutholakala inhlanganisela eyamukelekayo ehlangabezana nezindinganiso zekhwalithi ephansi kanye nokuvumelana.

Okulandelayo kuza isigaba sokuhlola kanye nokulungisa, lapho Amasethi edatha amakhulu futhi ahlukahlukene asetshenziswa ukwenza izivivinyo zokuqhathanisa ezinzimaKuphela uma ukugeleza kubonisa ukusebenza okuhambisanayo okuhambisana nezindinganiso ezichaziwe lapho kubhekwa khona ukuthi kukulungele isinyathelo esilandelayo.

Ekugcineni, esigabeni sokuqaliswa, ukugeleza kwenziwa ngcono ukuze kusebenze kahle futhi kusetshenziswe ekukhiqizeni. kufaka phakathi izinketho zokusebenzisa i-A/B, ukuqapha, ukuqoqwa kwempendulo yomsebenzisi, kanye nemijikelezo yokuthuthukisa eqhubekayoAkukho lutho olubekwe eceleni: ukugeleza kuyaqhubeka nokulungiswa ngokusekelwe kulokho okubonwayo ekusetshenzisweni kwangempela.

Le ndlela ipakishwe kuthempulethi yokugcina ye-GenAIOps, ene-code-first, amapayipi akhiwe ngaphambilini, kanye Amathuluzi okusebenzisa endawo nasekelwe efwini okuthuthukisa, ukuhlola, nokufaka izinhlelo zokusebenza ezisekelwe e-LLM ngaphandle kokuvuselela isondo kuphrojekthi ngayinye.

Ukuhlanganiswa ne-Azure DevOps: izindawo zokugcina, amapayipi, kanye nokuqinisekiswa

Ukuletha i-GenAIOps kusuka ku-theory iye enhlanganweni yangempela, ukuyihlanganisa ne-Azure DevOps kubalulekile. Ithempulethi evamile iqala ngo indawo yokugcina izinto e-Azure Repos enamagatsha amabili amakhulu, i-main kanye ne-development, okubonisa izindawo ezahlukene kanye namasu okukhuthaza ikhodi.

Indawo yokugcina isampula iqoshwe kusuka ku-GitHub, ehlotshaniswa ne-Azure Repos, kanye Ngokuvamile sisebenza ngokudala amagatsha ezici kusukela ekuthuthukisweni.Izinguquko zithunyelwa ngezicelo zokudonsa, eziqala ngokuzenzakalelayo ukuqinisekiswa kanye nezindlela zokuhlola.

Ukuze i-Azure DevOps isebenzisane ne-Azure Machine Learning nezinye izinsizakalo, ilungiselelwe inhlangano yesevisi e-Azure njengobunikazi bobuchwephesheLobu bunikazi busetshenziswa ekuxhumaneni kwesevisi ye-Azure DevOps, ngakho-ke amapayipi aqinisekiswa ngaphandle kokuveza okhiye embhalweni ocacile.

Ngokuvamile, le nhlangano inezimvume zomnikazi ekubhaliseni kwe-ML noma kumthombo osebenzayo, ukuze Amapayipi angahlinzeka ngama-endpoints, abhalise amamodeli, futhi abuyekeze izinqubomgomo ezitolo ezibalulekileUma ufuna ukuqinisa ukuphepha, ungalungisa indima ibe yi-Contributor ngokushintsha izinyathelo ze-YAML ezisingatha izimvume.

Ngaphezu kwalokho, iqembu lezinto eziguquguqukayo lenziwa ku-Azure DevOps ezithi Igcina idatha ebucayi njengegama lokuxhumeka kwesevisi noma izihlonzi zezinsiza.Lezi ziguquguquko zivezwa njengendawo ezungezile emipayipini, zigwema ukufaka amakhodi aqinile kulwazi olubalulekile kukhodi.

Ukulungiselela izindawo zokugcina zasendaweni nezikude kukuvumela ukuthi Igatsha lokuthuthukisa livikelwe ngezinqubomgomo zegatsha ezidinga ukuthi kwenziwe ipayipi lesicelo sokudonsa ngaphambi kokuvumela ukuhlanganiswa. Leli payipi liphatha ukuqinisekiswa kokwakha kanye nokugeleza kokuhlola, okuvimbela izinguquko eziphukile ukuthi zingeniswe.

Uma ikhodi isingena ekuthuthukisweni, ipayipi lokuthuthukiswa liyaqala ukusebenza Kuhlanganisa izigaba eziphelele ze-CI kanye ne-CD: ukusebenzisa izivivinyo nokuhlola, ukurekhoda ukugeleza kurejista yemodeli ye-Azure ML, ukusebenzisa ama-endpoints kanye nokuhlolwa komusi, kanye nokuhlanganisa kuma-endpoints asanda kudalwa.

Iphethini efanayo iphindaphindwa kulo lonke uhlobo noma igatsha lokukhishwa, elixhunywe ezindaweni zokukhiqiza. Lapho, Amapayipi e-CI/CD okukhiqiza aphinda umjikelezo wokuhlola, ukuhlola, kanye nokusetshenziswakodwa kudatha yengqalasizinda kanye nezinga lokukhiqiza, ngokulawula okukhulu kanye nokubuyekezwa okwengeziwe ngesandla uma kudingeka.

Imininingwane ebalulekile "ukubuyekezwa kwe-human loop" okufakwe kulezi zindlela: Ngemva kwesigaba se-CI, i-CD ihlala ikhiyiwe kuze kube yilapho umuntu eyivuma ngesandla. Ukuqhubeka kuvela ku-interface ye-Azure Pipelines. Uma kungavunyelwanga phakathi nesikhathi esithile (isibonelo, imizuzu engama-60), ukusebenza kuyenqatshwa.

Ukuqaliswa kwendawo kanye nokuxhumana nabahlinzeki be-LLM

Akuyona yonke into ejikeleza emipayipini: I-GenAIOps nayo iyasekela ukwenziwa kwendawo ukuze kwenziwe ukuhlolwa okusheshayoUngakopisha indawo yokugcina ithempulethi, udale ifayela le-.env kusiqondisi sempande, bese uchaza ukuxhumana ne-Azure OpenAI noma ezinye iziphetho ezihambisanayo ngaphakathi kwayo.

Lokhu kuxhumana kufaka phakathi amapharamitha afana ne-api_key, i-api_base, i-api_type, kanye ne-api_version, kanye Zibizwa ngamagama ngaphakathi kwezindlela (isibonelo, uxhumano olubizwa ngokuthi "aoai" olunenguqulo ethile ye-API). Ngale ndlela, ukugeleza okufanayo kungenziwa endaweni kanye nasefwini ngaphandle kokushintsha ikhodi.

Ukuze usebenzise le modi, mane nje dala indawo ebonakalayo noma i-conda bese ufaka ukuncika okudingekayo (i-promptflow, amathuluzi e-promptflow, i-promptflow-sdk, i-openai, i-jinja2, i-python-dotenv, njll.). Ukusuka lapho, ungabhala izikripthi zokuhlola kufolda yokusebenza yendawo bese usebenzisa izivivinyo ku-flows echaziwe.

Leli fu/indawo lihlangana kahle kakhulu nomqondo we-DevOps ovuthiwe: Ihlolwa ngezinga elincane endaweni, iqinisekiswe ngokusemthethweni emigwaqweni, bese ikhushulelwa ezindaweni eziphakeme ngokulawula nokuhlola.Konke kuhunyushwe ku-Git futhi kuxhunywe ku-Azure DevOps.

Amathuluzi ajwayelekile ohlelweni lwe-DevOps olune-AI kanye ne-LLMOps

Ngaphandle kokunikezwa okukhethekile kwe-Azure, uhlelo lwesimanje lwe-DevOps olune-AI kanye ne-LLMOps luvame ukuthembela ku- isethi yamathuluzi ahlanganisa i-ChatOps, ukuhlelwa kwemodeli, ukuqapha, kanye nokubuka.

Kusendlalelo se-ChatOps, kuvamile ukuhlanganisa USlack usebenzisa ama-bot afana ne-HubotI-Microsoft isebenzisana nama-ejenti asekelwe kuma-Power Virtual Agents, noma i-Discord kanye nezinhlaka ezifana ne-Botpress noma i-Rasa ukwakha abasizi abangokwezifiso abaxhumana namapayipi, izinhlelo zokuqapha, kanye nezinsizakalo zangaphakathi.

Ezindizeni ze-LLMOps/MLOps, zivame kakhulu amapulatifomu afana ne-Kubeflow kanye ne-MLflow ukuphatha amapayipi, amarekhodi amamodeli kanye nokuhlolwa, kanye namathuluzi athile njenge-Weights & Biases (W&B) yokulandelela okuthuthukisiwe kwe-metric, ukuqhathanisa okusebenzayo noma ukubonwa okujulile.

Kuhlelo lokusebenza lokwakha ku-LLM, kuvamile ukusetshenziswa izinhlaka ezifana neLangChain noma imitapo yolwazi yohlobo lwe-OpenLLMLezi zixazululo zenza kube lula ukuhlanganiswa kwamaketanga okulandelela, izixhumi kudatha yangaphandle, amathuluzi, kanye nama-ejenti anezinyathelo eziningi. Ngesikhathi esifanayo, kuvela izixazululo zokubonwa kwe-LLM ethile, okuvumela ukuqapha izimpendulo, izimpendulo, izindleko, kanye nekhwalithi.

Ngokuhlanganiswa ne-DevOps yakudala, amathuluzi afana ne-Jenkins noma i-GitLab CI ahlala efanele engxenyeni ye-CI/CD, I-Kubernetes ne-ArgoCD yokusetshenziswa okuqhubekayo kwe-cloud-nativekanye nezinqwaba zokubonwa ezifana ne-Prometheus, i-Grafana, ne-Loki zama-metric, amadeshibhodi, nama-log.

Izinselele, imikhawulo kanye nokwamukelwa okuqhubekayo

Konke lokhu kusetshenziswa kwemikhuba namathuluzi akuveli mahhala. Ubunzima bokuphatha izikhuthazo, izinguqulo zamamodeli, kanye nezinhlobo zokugeleza kukhulu kakhulu, ikakhulukazi lapho amaqembu amaningi esebenza ngesikhathi esisodwa —isimo lapho kunconywa khona ukufaka isicelo amasu afana ne-GitOps ukuhlanganisa izinguquko kanye nokusetshenziswa.

Ngaphezu kwalokho, ama-robot e-ChatOps kanye nama-LLM ngokwawo anekhono lokwenza okuthile Zethula izingozi ezinkulu zokuphepha uma benezimvume eziningi ezindaweni zokukhiqiza noma uma izindawo zokudalulwa kwedatha zingalawulwa kahle.

Okungeziwe kulokhu yi- ukuncika kumamodeli omthombo ovulekile anamalayisense abucayi noma ama-API ezentengiselwano okungashintsha izimo, amanani, noma imikhawulo. Futhi, okwenza izinto zibe zimbi nakakhulu, ukuhlolwa okuqinile kwama-LLM ekukhiqizweni kusalokhu kuyindawo evulekile, kunemibuzo eminingi engakaphendulwa.

Ngakho-ke, kunengqondo ukubhekana nokwamukelwa kwe-LLMOps kanye ne-ChatOps ngaphakathi kwe-DevOps ngendlela eqhubekayo nelawulwayo, ngokuqala ngokwenza imisebenzi ephindaphindwayo ngokuzenzakalela ngama-bot alula (ukuqalisa kabusha, imibuzo yelogi, ukumaka kokwakha, njll.).

Kamuva, zingafakwa I-LLM yemisebenzi yokusekela, ukuhlukaniswa kwezigameko, noma usizo lokulungisa amaphuthaIsibonelo, ngokuchaza amaphutha ngokusekelwe kumarekhodi noma ngokuphakamisa izindlela zokunciphisa ngokusekelwe kumadokhumenti angaphakathi.

Uma ukusebenza kwe-ML yakudala sekuzinzile, sekuyisikhathi sokuthi ikheli le-LLMOp elinamamodeli olimi akhethekile kuma-domain afana nesevisi yamakhasimende, i-DevSecOps noma i-QA, sisebenzisa konke okufundwe ezigabeni zangaphambilini.

Indawo lapho yonke le mikhuba ikhomba khona indawo yobunjiniyela exoxayo, ebikezelayo, futhi ekhula ngokuzimelalapho iningi lentuthuko kanye nokusebenza kuvezwa ngolimi lwemvelo futhi i-AI isiza ekwenzeni izinqumo ezisebenzayo mayelana nokusetshenziswa, ukukhuliswa noma ukubuyiselwa emuva.

Njengoba le nkinga ikhona—ama-DevOps, ama-ChatOps, ama-MLOps, ama-GenAIOps, nama-LLMOps—izinhlangano sezinayo uhlaka oluqinile lokwakha nokusekela izinhlelo ezisekelwe ku-LLM eziletha ngempela inaniUkugcina ukulawula ikhwalithi, izindleko, ukuphepha, kanye nokuhambisana nebhizinisi, esikhundleni sokuhlala nezibonelo ezilula noma izivivinyo ezihlukanisiwe eziwa ngokushesha nje lapho zifika ekukhiqizweni.