I-OpenAI ngoku ivumela ukulungelelanisa inkqubo yokuvelisa umbhalo we-GPT-3

I-OpenAI, ilabhoratri esekwe eSan Francisco, eKhalifoniya ephuhlisa ubugcisa bobuchule bokwenziwa obuquka imifuziselo yolwimi olukhulu, ubhengeze ukukwazi ukwenza iinguqulelo eziqhelekileyo ze-GPT-3, imodeli ekwazi ukuvelisa ikhowudi yohlobo lomntu kwisicatshulwa kunye nentetho.

Ngayo ukususela ngoku ukuya phambili ababhekisi phambili banokusebenzisa ulungiso olucikizekileyo ukwenza imifuziselo ye GPT-3 elungiselelwe umxholo othile wezicelo zakho kunye neenkonzo, okukhokelela kwiziphumo ezikumgangatho ophezulu kuyo yonke imisebenzi kunye nomthwalo onzima, kuxhomekeke kwinkampani.

Kwabo bangayazi i-GPT-3, kufuneka uyazi ukuba oku yimodeli yolwimi oluzenzekelayo esebenzisa ukufunda nzulu ukuvelisa izicatshulwa ezifana nomntu.

Nguwo imodeli yokuqikelela ulwimi yoluhlu lwesizukulwana sesithathu lwe-GPT-n oludalwe yi-OpenAI, ilabhoratri yophando yobukrelekrele esekwe eSan Francisco eyenziwe yinkampani eyenza inzuzo i-OpenAI LP kunye nenkampani engumzali, inkampani engenzi ngeniso i-OpenAI Inc.

Kuwo nawuphi na umyalezo obhaliweyo, onjengesivakalisi, i-GPT-3 ibuyisela umbhalo owongezelelweyo ngolwimi lwendalo.

Abaphuhlisi Bangakwazi 'iprogram' ye-GPT-3 ngokukubonisa imizekelo embalwa okanye 'i-prompts'.

"Siyile i-API ukuba ibe lula kumntu wonke ukuba ayisebenzise kwaye ibe bhetyebhetye ngokwaneleyo ukwenza amaqela okufunda ngomatshini abe nemveliso ngakumbi," utshilo u-OpenAI ngasekupheleni kukaMatshi.

Ngeli xesha, izicelo ezingaphezu kwama-300 basebenzisa i-GPT-3 kwiindidi ezahlukeneyo kunye namashishini, ukusuka kwimveliso kunye nemfundo nkqu ubuchule kunye nemidlalo.

La umthamo omtsha wokucokisa kwi-GPT-3 setting ivumela abathengi ukuba baqeqeshe i-GPT-3 ukuqonda ipateni ethile yomthwalo womsebenzi njengokuveliswa komxholo, ukuhlelwa kunye nokudityaniswa kwesicatshulwa ngaphakathi kwemida yendawo ethile.

Umboneleli osebenzayo usebenzisa i-GPT-3 ukunceda iinkampani ukuba zisebenzise ingxelo yabathengi. Ukusebenzisa idatha engacwangciswanga, inkqubo inokuvelisa iingxelo ezishwankathela ingxelo yabathengi kunye nokusebenzisana. Ngokwenza ngokwezifiso i-GPT-3, i-Viable yayiza kukwazi ukwandisa ukuchaneka kweengxelo zayo ukusuka kwi-66% ukuya kwi-90%.

Okufanayo kuya kuMgcini weRhafu, isixhobo esenza lula ukuziphendulela ngokuzicwangcisa ngokuzenzekelayo kunye nokukhupha idatha yokuhlawula umvuzo weengxelo zerhafu, kwibhanki okanye kwi-akhawunti yokuhlawula. Irhafu yomgcini usebenzisa i-GPT-3 ukutolika idatha yengxelo yebhanki ukufumana iindleko ezinokutsalwa irhafu. Inkampani iyaqhubeka nokucokisa i-GPT-3 ngedatha entsha ngeveki nganye ngokusekelwe kwintsebenzo yehlabathi yangempela yemveliso yayo, igxininise kwimizekelo apho imodeli iwele ngaphantsi komda othile wokusebenza.

Los abaphuhlisi bongeza malunga nama-500 iisampuli ezintsha ngeveki ukucokisa imodeli. Umgcini weRhafu uthi inkqubo yokulungelelanisa ivelisa i-1% yokuphucula ukusuka ngeveki ukuya ngeveki.

«Inye into ebesilumke ngayo kwaye sinyanzelise kuphuhliso lwethu lwale API kukwenza ukuba ifikeleleke kubaphuhlisi abangenayo imvelaphi yokufunda koomatshini, utshilo uRachel Lim, ilungu labasebenzi bobugcisa be-OpenAI. "Indlela ephuma ngayo kukuba unokwenza itemplate ye-GPT-3 ngokwezifiso usebenzisa i-invocation yomgca womyalelo. [Sinethemba] ukuba ngenxa yokufikeleleka kwayo, sinokufikelela kwiseti yabasebenzisi abahlukeneyo abanokuza nezona ngxaki zabo zahlukeneyo kubuchwephesha. "

U-Lim uthi amandla okuphucula i-GPT-3 angakhokelela ekongiweni kweendleko njengoko abathengi banokulindela ukuphindaphinda okuphezulu kweziphumo zomgangatho ongcono ukusuka kwiimodeli ezifakwe ngokuchanekileyo xa kuthelekiswa nemodeli eqhelekileyo ye-GPT-3. inani leethokheni, okanye amagama, avelisiweyo yimifuziselo.)

Ngelixa i-OpenAI ineprimiyamu kwiimodeli ezisulungekileyo, i-Lim ithi uninzi lweemodeli ezisulungekileyo zifuna ukuhanjiswa okufutshane kunye namathokheni ambalwa, anokonga imali.

I-GPT-3 API ifumaneke esidlangalaleni ukususela ngo-2020. Unyaka ngaphambi kokuba iqaliswe, abaqulunqi bayo bagqibe ekubeni bangenzi esidlangalaleni umsebenzi wophuhliso lwenguqu yangaphambili, i-GPT-2, ngokuqwalasela ukuba le nkqubo i-doped ngokufunda ngomatshini inokujika. ibe yingozi xa ithe yawela ezandleni zabantu abakhohlakeleyo.


Umxholo wenqaku uyabambelela kwimigaqo yethu imigaqo yokuziphatha yokuhlela. Ukuxela impazamo cofa apha.

Yiba ngowokuqala ukuphawula

Shiya uluvo lwakho

Idilesi yakho ye email aziyi kupapashwa.

*

*

  1. Uxanduva lwedatha: UMiguel Ángel Gatón
  2. Injongo yedatha: Ulawulo lwe-SPAM, ulawulo lwezimvo.
  3. Umthetho: Imvume yakho
  4. Unxibelelwano lwedatha: Idatha ayizukuhanjiswa kubantu besithathu ngaphandle koxanduva lomthetho.
  5. Ukugcinwa kweenkcukacha
  6. Amalungelo: Ngalo naliphi na ixesha unganciphisa, uphinde uphinde ucime ulwazi lwakho.