Ukufunda ngomshini
Ithegi yethuluzi le-AIEZZ
Ubheka imikhiqizo ye-AI enethegi ethi “Ukufunda ngomshini”, enemiphumela efanayo engu-8.
ReplicateInkundla yokufunda nge-AII-Replicate iyinkundla ye-AI esebenza efwini eqondiswe kubathuthukisi, ehlinzeka ngamamodeli omthombo ovulekile emikhakheni efana nezithombe, amavidiyo, imibhalo nomculo. Abasebenzisi bangabuka indlela yokusebenzisa ekhasini lemodeli, baxhume amamodeli ezinhlelweni besebenzisa i-API ejwayelekile, noma basebenzise amathuluzi enkundla ukulungisa amamodeli noma ukuthumela amamodeli abo. Ilungele abathuthukisi bezinhlelo zokusebenza, amaqembu aqalayo, abacwaningi be-AI namaqembu ezinkampani adinga ukuhlola izixazululo zamamodeli, ukuze bakhe ama-prototype ngokushesha noma bahlanganise imisebenzi ye-AI ekhiqizayo. Inkundla ikhokhisa ngokwesikhathi sangempela sokusebenza futhi isekela ukulungiswa kwezinsiza ngokuya ngenani lezicelo. Kumele kuqashelwe ukuthi amamodeli ahlukene angahluka ngekhwalithi, isikhathi sokuphendula, ukusetshenziswa kwezinsiza nemibandela yokusebenzisa. Ngaphambi kokusetshenziswa ngokusemthethweni, kusadingeka ukuhlola ngokwezidingo zebhizinisi nokulinganisa izindleko.028.9
kaggleInkundla yokufunda nge-AIi-kaggle iyinkundla eyakhelwe ukufunda nokusebenzisa isayensi yedatha nokufunda komshini, ihlanganisa imincintiswano, amasethi edatha avulekile, ama-Notebook aku-inthanethi, izifundo nokuxhumana komphakathi. Abasebenzisi bangakhetha izifundo ukuze bafunde izisekelo ze-Python, R, SQL nokufunda komshini, bese besebenzisa amasethi edatha nendawo yokuhlela eku-inthanethi ukwenza izivivinyo. Bangabamba iqhaza nasemincintiswaneni ukuze bahlole amakhono abo okwakha amamodeli. Le nkundla iqondiswe kwabaqalayo besayensi yedatha, abafundi, onjiniyela bama-algorithm namaqembu adinga ukwenza ukuhlaziya kwedatha. Le nkundla iphinde inikeze okuqukethwe noma izinsiza zamahhala nezikhokhelwayo, kodwa amandla ekhompyutha asefuzeni, izifundo nemithetho yemincintiswano kungase kube nemibandela yokusetshenziswa; imininingwane kufanele iqinisekiswe ekhasini lenkundla.038.9
ML For BeginnersInkundla yokufunda nge-AII-ML For Beginners iyisifundo samahhala esinomthombo ovulekile sokwethulwa ekufundeni ngomshini, esethulwe yi-Microsoft futhi esisingathwe ku-GitHub. Senzelwe abaqalayo, abafundi, abantu abashintsha imisebenzi kanye nabathuthukisi abafuna ukuqonda ukufunda ngomshini ngendlela ehlelekile. Isifundo sihlelwe saba amasonto angu-12 nezifundo ezingu-26, sigxile ezihlokweni ezifana nokubikezela, ukuhlukanisa, ukuqoqa amaqembu kanye nokucubungula ulimi lwemvelo. Sihlanganisa amaphrojekthi angokoqobo, imibuzo yokuzihlola ngaphambi kwesifundo, izincazelo zezifundo, izibonelo zekhodi ze-Jupyter Notebook, imisebenzi yangemva kwesifundo namavidiyo okufunda. Sisebenzisa kakhulu i-Python kanye namathuluzi afana ne-Scikit-learn, Pandas ne-Matplotlib, futhi sifanele ukuzilolonga kancane kancane ngokulandela indlela ebekiwe. Qaphela ukuthi sigxile ezisekelweni zokufunda ngomshini kwakudala, ngakho asikwazi ukuthatha indawo yokuqeqeshwa kokufunda okujulile noma kochwepheshe okuthuthuke kakhulu.038.6
Udacity AI AcademyInkundla yokufunda nge-AII-Udacity AI Academy iyinkundla yokufunda egxile emsebenzini, eyenzelwe abaqalayo ku-AI, abantu abaguqula imisebenzi, onjiniyela abasebenzayo nabafundi. Ihlinzeka ngezifundo emikhakheni efana nokufunda komshini, ukufunda okujulile, ukubona kwekhompyutha, ukucubungula ulimi lwemvelo, ukufunda ngokuqinisa kanye ne-AI ekhiqizayo. Inkundla igxile ezinhlelweni ze-Nanodegree nasekufundeni okuqhutshwa ngamaphrojekthi, lapho abafundi beqedela imisebenzi yokuhlela nokuhlaziya esekelwe ezimweni zangempela, kancane kancane bahlanganise amaphrojekthi abangawakhombisa. Ezinye izinsizakalo zokufunda ezikhokhelwayo zihlanganisa ukubuyekezwa kwekhodi ngumeluleki, ukuqondiswa kwamaphrojekthi nokwesekwa komsebenzi, futhi zilungele abantu abafuna ukuthuthukisa amakhono abo ngendlela ehlelekile noma balungiselele ukungena emisebenzini ehlobene. Qaphela ukuthi izinsizakalo ezifana nomeluleki nosizo lokufuna umsebenzi zitholakala kubafundi abakhokhelayo, futhi abafundi kusadingeka batshale isikhathi baqede amaphrojekthi ngokuya ngesisekelo sabo.038.4
Google AI StudioUkuhlela nge-AII-Google AI Studio iyinkundla yokuthuthukisa i-AI esebenza esipheqululini, eqondiswe kubathuthukisi, abafundi, abacwaningi, abathanda i-AI namaqembu adinga ukuhlola imibono. Abasebenzisi bangakhetha izindlela zokwazisa ezikhululekile, ezihlelekile noma zengxoxo ekhasini, bafake umbhalo, izithombe, umsindo namavidiyo, bese benza ama-prototype futhi balungise amamodeli e-Gemini, kuhlanganise nokulungisa amapharamitha afana ne-Temperature ne-Top-K. Ngemva kokuqeda ukuhlola, inkundla ingakhiqiza ikhodi yokubiza i-API ngezilimi ezifana ne-Python ne-Node.js, futhi isekela ukuthumela ama-prototype ku-Google Cloud nokukhiqiza izixhumanisi zokwabelana. Le sevisi iyatholakala mahhala, kodwa imiphumela yamamodeli nekhodi ekhiqiziwe kusafanele kuhlolwe futhi kubuyekezwe ngokuvumelana nephrojekthi ethile.189.2
WandbAmathuluzi Okuhlela nge-AII-Wandb iyithuluzi lokulandelela izivivinyo nokubona ngeso lengqondo elenzelwe abacwaningi be-AI, onjiniyela bama-algorithm namathimba okufunda komshini. Lisetshenziswa kakhulu ukuqopha ngokuhlanganyela inqubo yokuqeqesha, ukuqhathanisa imiphumela yezivivinyo nokusiza ukubambisana kwethimba. Lapho lisetshenziswa, onjiniyela baxhuma i-Wandb kukhodi yokuqeqesha ukuze baqophe ama-hyperparameter, izinkomba zamamodeli, izinguqulo zekhodi kanye nolwazi lwesistimu olufana ne-CPU ne-GPU, bese bebuka futhi baqhathanise izivivinyo kudeshibhodi eku-inthanethi. Inkundla iphinde inikeze ukusesha kwama-hyperparameter kanye namandla okuphatha izinguqulo zamamodeli namasethi edatha, futhi ifanele izinqubo zokuthuthukisa ezidinga ukuhlela imijikelezo eminingi yezivivinyo nokuhlaziya ukusebenza kwamamodeli. Ulwazi olutholakalayo lubonisa ukuthi inikeza izinhlelo zamahhala nezikhokhelwayo. Ukusetshenziswa kwangempela kusadinga ukuqedwa kokuxhunywa kwekhodi, futhi indlela yokuqopha efwini kufanele ihlolwe ngokuhambisana nezidingo zokuphathwa kwedatha zethimba.047.8
PyTorchAmathuluzi Okuhlela nge-AII-PyTorch iwuhlaka oluvulekile lokufunda komshini olwasungulwa ithimba locwaningo le-Meta AI, futhi manje luphethwe yi-Linux Foundation. Lusebenza kakhulu kubacwaningi bezobuhlakani bokwenziwa, onjiniyela bokufunda komshini, ososayensi bedatha nabafundi. Abasebenzisi bavame ukusebenzisa i-Python ukuchaza amamodeli nezinqubo zokubala eziya phambili, bese besebenzisa ukubala kwama-tensor, i-automatic differentiation namamojuli amanethiwekhi ezinzwa ukuze baqeqeshe, kanye nemitapo yolwazi ehlobene ukuphatha imisebenzi yombono, umsindo noma umbhalo. Igrafu yokubala eyakhiwa ngesikhathi sokusebenza yenza ukuhlola nokulungisa kube lula, futhi isekela ukuqhubekisela phambili i-prototype yocwaningo iye ekusetshenzisweni kwemodeli. Ukusetshenziswa kwangempela kusadinga ulwazi lwe-Python, ukubala kwama-tensor nezisekelo zokufunda okujulile; lapho ikhodi yokuhlola idluliselwa endaweni yokukhiqiza, kudingeka futhi uphathe ngokwakho ukuthunyelwa kobunjiniyela nokumiswa kwezinsiza.057.6
Hugging FaceIbhokisi lamathuluzi e-AII-Hugging Face iyinkundla yokufunda komshini eyenzelwe abathuthukisi be-AI, ososayensi bedatha, abacwaningi nabafundi. Izinsiza zayo eziyinhloko zihlanganisa inqolobane yamamodeli, isikhungo samasethi edatha, umtapo we-Transformers kanye nama-Spaces. Abasebenzisi bangasesha imodeli noma isethi yedatha endaweni kuqala, bese beyibiza, benze izivivinyo futhi bayilungise kahle besebenzisa isixhumi esibonakalayo esifanayo; bangasebenzisa nama-Spaces ukubonisa izinhlelo ezisebenzisanayo. Uma kudingeka ukusatshalaliswa kokukhiqiza, bangafunda nangezinsizakalo zokuphetha ezisingathiwe. Le nkundla ilungele ukukhetha amamodeli, ukuqinisekisa ama-prototype, ukufundisa nocwaningo kanye nokuthuthukisa izinhlelo. Kufanele kuqashelwe ukuthi okuqukethwe kwenkundla kuvela kubanikeli abahlukene, ngakho ububanzi bokusetshenziswa, indawo encike kuyo kanye nemibandela yokusetshenziswa kwemodeli kungahluka; ngaphambi kokusetshenziswa kwangempela, kufanele kufundwe incazelo ehambisanayo.038.6Wonke amathuluzi analeli thegi ayaboniswa.