Usebenzisa i-NoteBooklm ukubhekana nemibuzo enzima: Ingxoxo ngobuchule, hhayi nzima

Isithombe nguMlobi

Okusobala Ukuqalisa

Phakamisa isandla sakho uma wake waqanda iqhwa ngesikhathi sengxoxo yezobuchwepheshe lapho umuntu oxoxisana naye ebuza khona, “ngihambe ngendlela yakho.” Iningi labazongenela ukhetho lifunda umbuzo, gxuma uqonde ngqo kwikhodi, futhi uthembele kwimemori yemisipha yokukhahlela.

Kuthiwani uma ungalayisha umbuzo wengxoxo kwithuluzi le-AI futhi uthole incazelo ye-podcast yekhodi yakho, imephu yomqondo ebonakalayo, ama-flashcards, kanye nemibuzo? Well, ungakwazi Incwajana.

Kulesi sihloko, sizoqala ukuxazulula “uhlelo lokuncoma” lwe-meta uqobo, bese sisebenzisa izici eziyisithupha ze-NotebookLm ukuqonda nokufunda kuyo ngobuchule, hhayi nzima.

Okusobala Hlangana ne-NotebookLM: Umsizi wokufunda we-AI-Powered Ai-Powered

I-NoteBookm ngumsizi wokutadisha we-AI we-AI. Iguqula indlela esifunda ngayo kwimininingwane ngokuhlanganisa izici eziningi ezinamandla ze-AI ezisebenzisa inqubo zisebenzisane kakhulu futhi zikwazi ukuzivumelanisa nezimo.

Kwenza njalo ngokuguqula amadokhumenti akho, izincwadi, noma ezinye ezinye izinto zibe ngamathuluzi okufunda asebenzayo. Ngokucacile, kuguqula izinto zakho zibe izingxoxo, amamephu abukwayo, kanye namamephu, futhi kusindisa amahora amaningi wokubuyekezwa okwenziwa ngesandla, okwenza kube lula ukugaya.

Lokhu kukusiza ukuthi uqonde futhi ugcine ulwazi ngokushesha. Ngamafuphi, i-noteookilm ikusiza ukuthi ufunde ngobuchule, hhayi nzima.

Manje ake sixazulule umbuzo we-meta interview, bese uhlola ukuthi i-norebookl ingasisiza kanjani ukuba sifunde kuleyo khodi efanayo.

Okusobala Inselelo: “Uhlelo lokuncoma” lweMeta

Uhlelo lokuncoma
Unikezwa uhlu lwabangani be-Facebook kanye nohlu lwamakhasi we-Facebook abasebenzisi abalandela abasebenzisi. Umsebenzi wakho ukudala uhlelo lokuncoma olusha lwe-Facebook. Kumsebenzisi ngamunye we-Facebook, thola amakhasi ukuthi lo msebenzisi angalilandeli kepha okungenani omunye wabangane babo asebeyenzayo. Khipha i-ID yomsebenzisi kanye ne-ID yekhasi okufanele inconyelwe kulo msebenzisi.

Umbuzo uyatholakala Stratascratch. Kulo mbuzo, i-meta yasicela ukuba sakhe uhlelo lokuncoma olubonisa amakhasi e-Facebook umsebenzisi akasalilandeli okwamanje, kepha okungenani omunye wabangane babo enza.

// Ukuqonda idatha ngemuva kwenkinga

Sinezinto ezimbili ezihlukile, user_friends na- user_pages.

Owokuqala okhombisa ukuxhumana komsebenzisi ngamunye (obani abangane abane), ngenkathi user_pages Ibala ukuthi yimaphi amakhasi umsebenzisi ngamunye asevele elandelayo.

Ngokubambisana, basisiza ukuthi sithole izincomo zekhasi ezintsha ezisuselwa kumsebenzi wabangane abahlanganyeli.

Ake sibheke kuqala user_friends dataset kuqala.

Ake ubuke kuqala idatha yesibili manje, user_pages:

// Isixazululo Sesinyathelo Ngesinyathelo: Ukwakha Injini Yokuncoma

Umgomo ukuncoma amakhasi we-Facebook ukuthi umsebenzisi angakasilandeli okwamanje, kepha okungenani omunye wabangane babo asevele elandelayo.

Okokuqala, sixhumanisa umsebenzisi ngamunye namakhasi abangane babo abakulandelayo. Lokhu kusisiza ukuthi sibone ukuthi yimaphi amakhasi alandelwa yinethiwekhi yomsebenzisi kanye nezincomo ezingaba khona.

friends_pages = users_friends.merge(users_pages, left_on='friend_id', right_on='user_id')
friends_pages = friends_pages[['user_id_x', 'page_id']]

Manje sidinga ukukhipha amakhasi abasebenzisi asevele basilandelayo. Senza lokhu ngokuhlanganisa futhi nokwangempela users_pages Ithebula nokugcina amakhasi kuphela angekho ohlwini lomsebenzisi.

comparison = friends_pages.merge(
    users_pages,
    how='left',
    left_on=['user_id_x', 'page_id'],
    right_on=['user_id', 'page_id']
)
result = comparison[comparison['user_id'].isna()][['user_id_x', 'page_id']]

Ekugcineni, sisusa noma yikuphi okuphindwe kabili ukuze sigweme ukuncoma ekhasini elifanayo kaninginingi futhi layiqamba kabusha ikholomu yokucaca.

result = result.drop_duplicates()
result = result.rename(columns={'user_id_x': 'user_id'})

Nayi imigqa embalwa yokuqala yokuphuma okulindelekile.

Kuze kube manje, sithembele kumasu endabuko: ukufaka amakhodi, ukuhlanganisa, nokuhumusha konke ngokwethu.

Manje, i-seteBoodlulm yakwaGoogle isivumela ukuthi sithathe inqubo yokufunda sibe isinyathelo futhi.

Okusobala Ukusuka Ekuxazululeni Ukufunda: Faka i-Notebooklm

Kulesi sigaba, sizokhombisa ukunikeza i-Notebookm ngemininingwane yombuzo we-meta wenhlosi, bese uhlola izici zayo eziyisithupha ezisebenzayo ezikusiza ukuba ufunde ngobunono.

// Isinyathelo 1: Ukwakha incwadi entsha futhi ungeze idatha yakho

Ngaphambi kokuthi i-pateBooklm ingasisiza, kudingeka sisinikeze imininingwane ngombuzo wethu wenhlolokhono. Lokhu kuqala ngokwakha incwadi entsha. Ikhanda ku Isayithi le-Notebooklm bese uchofoza u- “Dala incwadi entsha.”

Izokubuza ngomthombo.

Njengoba ubona esikrinini esingezansi, izinketho ezahlukene ziyatholakala.

• I-Google Drayivu
• Iwebhusayithi
• I-YouTube
• Namathisela umbhalo

Masisebenzise inketho ethi “Paste Pest” bese unamathisela imetadata yombuzo we-meta wenhlolokhono. Nalu ulwazi oluzofaka khona ukusebenzisa le fomethi ephindeka kabusha.

Here is the question.
[paste question here]
It uses two datasets.
[paste dataset information here]
Here is the python solution.
[paste python solution here.]

Okulandelayo, chofoza ku- “Faka”, kuboniswe ngezansi.

// Ukuhlola Isitudiyo sokufunda esisebenzayo

Ngezansi kwesikrini esivula ngemuva kokushaya “Faka.” Ngakwesobunxele, singabona imithombo. Maphakathi, sine-LLM eyayiqeqeshelwa imininingwane yethu. Ngakwesokudla, sine- “studio” equkethe izinhlobo eziyisithupha zokufunda ezihlukile, esizozihlola ngamunye ngamunye.

Ase sichofoze “I-Python sixakeke kanjani inzuzo yedatha yomngane nekhasi lezincomo zezincomo”, ezingaphansi kombhalo wokufaka phakathi. Lo mbuzo wakhiqizwa yi-NoteBooklm.

Nansi impendulo.

Njengoba ukwazi ukubona, kuchaza wonke umqondo. Ungagcina le nothi ngokuchofoza “Gcina inothi” ekugcineni. Amanothi azovela ngaphansi kwesigaba se- “Studio” kanjena.

Ngakho-ke, uma unemibuzo ehlobene nalo mbuzo wengxoxo, ungasebenzisa i-LLM phakathi kwesikrini esedlule ukuthola impendulo. Ake sihlole izici eziyisithupha ze- “Studio” eziyisithupha ezenza izinto zithakazelise.

// Ukuthola izici zokufunda eziyisithupha ze-NoteBooklm

Kulesi sigaba, sizobona izici eziyisithupha zokufunda ze-Notebooklm, eziboniswe ngezansi, zisebenza.

Yilowo nalowo akusiza ukuthi uqonde umbuzo we-meta interview kusuka ku-angle ehlukile – ngokusebenzisa umsindo, ividiyo, okubukwayo, nokuzijwayeza okusebenzayo.

1. Khiqiza umbono olalelwayo
Chofoza u- “Audio Overview.”
Lapho wenza lokho, uzobona lesi saziso ngaphansi kwe- “studio,” futhi umbono wakho wokulalelwayo uzobe usulungile emizuzwini embalwa.

I-Notebooklm iguqula okuqukethwe kwakho okulayishiwe kungxoxo ye-podcast-style. Amagama amabili e-AI axoxa ngenkinga yakho sengathi ulalele umbukiso we-tech phrep show. Baphula uhlelo lokuncoma kaMeta, bachaze okunengqondo, futhi baqokomise amacala anqenqema.

Ungalanda le ngxoxo, futhi ngenhlanhla, ungayijoyina ngokuchofoza ku- “Interactive.” Isikrini esingezansi sizovula.

Futhi uma uchofoza ujoyine, umyalezo uzothi “Sawubona, othile ufuna ukujoyina,” ukuvumela ukuthi ujoyine ingxoxo. (Yebo, ungajoyina ingxoxo ngempela; Kuvuselelwa ngesikhathi sangempela nge-API-kuya-enkulumweni yenkulumo. Isici esizayo!)

2. Khiqiza ukuwohloka kwevidiyo
Buyela emuva kuPhaneli ethi “Studio” bese ushaya “ukubuka konke kwevidiyo.” I-Notebooklm idala ividiyo ebona idatha yakho futhi ichaze ikhambi. Kuthu, ividiyo iyimizuzu engu-6 nemizuzwana engama-21 ubude.

Kuvidiyo ekhiqizwayo, i-noteokbolm kuqala ichaza umqondo obanzi: ukuthi amapulatifomu ezokuxhumana anquma ukuncoma yini, ngaphambi kokuya enkingeni yethu ethile.

Okulandelayo, ividiyo ekhiqizwayo ixoxa ngenkinga yengxoxo ngokuyiqeda izingxenye, iqala ngalesi sinyathelo esengeziwe esiboniswe ngezansi.

Okulandelayo, kuqala ukuchaza ikhambi. Akumane nje ukuphinda okusesikrinini; Izincazelo zibuyela emcabangweni.

Futhi ekugcineni, kwehlisa ikhambi libe yizinyathelo futhi lilichaza lisebenzisa lezo zinyathelo.

3. Imephu I-Logic ngokubona
Phinda ubuye kuPhaneli ethi “Studio” bese uqhafaza “Imephu Yengqondo.”
I-Notebooklm ikhiqiza isihlahla esibonakalayo senkinga. Ake sibone eyodwa.

Ubona yonke inkemba yenkinga ngokubuka okukodwa: umgomo wokuncoma phezulu, ama-datasets adingekayo ezinyangeni eziphakathi nendawo, kanye nezinyathelo zokuxazulula ezansi. Ase sichofoze kwesinye sazo, ngokwesibonelo, “Isinyathelo 1: Khomba amakhasi wabangane (hlanganisa 1).”

Njengoba ukwazi ukubona, imephu yengqondo iyanda ngokuchaza isinyathelo 1. Ungayilanda, futhi.

4. Yakha imibiko
Buyela emuva kuphaneli yesitudiyo bese uqhafaza “Imibiko.” I-Notebooklm ikubuza ukuthi ukhethe ukuthi yiluphi uhlobo lombiko ofuna ukuwadala.

Ake sikhethe inkinga yokuhamba ngezinyawo. Kuzoqala ukukhiqiza umbiko.

Nakhu ukuhamba kwezinyathelo ngesinyathelo esichaza ukuthi ungawuxazulula kanjani lo mbuzo wengxoxo.

5. Khiqiza ama-flashcards
Phinda ubuye kuPhaneli ethi “Studio” bese uchofoza “Ama-Flashcards.” I-Notebooklm ikhiqiza amakhadi we-Q & a auto. Ake sibone enye yazo.

Futhi uma uchofoza impendulo, nangu umphumela.

Ase sichofoze “Chaza.” Isebenzisa ngokushesha ukuphendula lo mbuzo ngamamodeli we-Gemini.

Manje ake sibone imiphumela.

6. Dala umbuzo
Buyela emuva kuphaneli ye-studio bese uqhafaza kwimibuzo. I-Notebooklm yakha isivivinyo sokuzijwayeza. Ake sikwenze futhi sibone imibuzo yokuqala.

Ase sichofoze “ukusikisela.”

Impendulo manje isobala. Ngakho-ke, ake sikhethe futhi sibone ukuthi kwenzekani.

Uma usenombuzo, chofoza ku- “Chaza.” Kukuthatha uye esigabeni esiphakathi, lapho i-LLM eqeqeshelwe khona ngesisombululo sethu inikeza impendulo. Nakhu.

Okusobala Ukugcina

Ukulungiselela izingxoxo zezobuchwepheshe akusho ukuthi ugaya izinkinga zokugaya esolate. Kusho ukuqonda ngokujulile, ngeso lengqondo ngokusobala, futhi kuchaze ngokuzethemba. I-Notebooklm iguqula umbuzo owodwa wokuxoxisana kube yisipiliyoni sokufunda sezinzwa eziningi, kufaka phakathi umsindo, ividiyo, amamephu abukwayo, nemibiko ebhaliwe, nokukhumbula okusebenzayo.
Usuvele unamakhono okuxazulula izinkinga. Manje usunohlelo lokuwahlela, luziqede, futhi uzenzele ngaphansi kwengcindezi. Ngokuzayo lapho ubona umbuzo onzima we-SQL noma uPython, ungesabi; Yilayishe, uyihlole, uyiqonde kahle.

Nate rosidi Ungusosayensi wedatha kanye necebo lomkhiqizo. Ungumuntu futhi ongumsunguli wokufundisa ofundisayo, futhi umsunguli weStratascratch, isiteji esisiza ososayensi bedatha balungiselele izingxoxo zabo ngemibuzo yangempela yengxoxo evela ezinkampanini eziphezulu. UNate ubhala ezitayeleni zakamuva emakethe yezemisebenzi, unikeza izeluleko ezixoxiswayo, wabelana ngamaphrojekthi weSayensi yedatha, futhi amboze yonke into SQL.