Kedis ku xidhan hagaha · Dhexdhexaad Heerka

OpenAI Kediska Shaqsiga

Dib u eegis xogta tababarka-whisper, habdhaqanka luuqadaha badan, iyo hababka guul darrida ee la diiwaan galiyay adoon u malaynayn in qoraalladu ay had iyo jeer sax yihiin.

Wadooyinka haga ee la xidhiidhaOpenai Whisper
Su'aal 1 ee 8

Waa maxay ujeedada ugu weyn ee OpenAI's Whisper?

Horudhac su'aal kasta8 su'aalo

Jawaabaha iyo sharraxaadaha waxay soo baxayaan markaad qaadanayso kediska.

  1. Waa maxay ujeedada ugu weyn ee OpenAI's Whisper?
    • Aqoonsiga tooska ah ee hadalka: qorista iyo tarjumida maqalka
    • Ka soo saarida sawirada qoraalka
    • Qorista koodka kombiyuutarka
    • Abuuritaanka animations 3D
  2. What was true of the language balance in Whisper’s training audio?
    • Every language contributed the same number of hours
    • The dataset was English-heavy, with much less audio for most languages
    • The dataset contained only English audio
    • Most data was studio-recorded speech in one language
  3. What does Whisper use to condition the decoder on the requested task and language?
    • A separate camera feed
    • A hand-written transcript of the answer
    • Special tokens in the decoder prompt
    • A unique model checkpoint for every utterance
  4. Awooddee ayuu Whisper AAN si asal ahaan u bixin?
    • U turjumida hadalka luqadaha kale una turjumaya Ingiriis
    • Ogaanshaha luqadda lagu hadlo
    • Ku-daridda waqti-qaadaha qoraalka qoraalka
    • Ka abuurista muuqaal macquul ah oo degdeg ah
  5. Which long-form transcription problems does OpenAI’s paper describe?
    • Repetition loops and missing words near segment edges
    • Perfect diarization of every speaker
    • Automatic correction of every proper noun
    • Guaranteed removal of background music
  6. Waa maxay daciifnimada la yaqaan ee Whisper in la daawado?
    • Ma farsamayn karo wax maqal Ingiriisi ah
    • Waxay mararka qaarkood 'dhalaali kartaa' qoraalka aan lagu hadlin
    • Kaliya waxay soo saartaa maqal, waligeed qoraal
    • Way diiday inay ku darto xarakaynta
  7. Approximately what share of Whisper’s training data was English audio paired with English transcripts?
    • About 18%
    • Ilaa 1%
    • About 65%
    • About 100%
  8. What use does OpenAI caution against when transcription errors could have serious consequences?
    • Noncommercial experimentation with public samples
    • Ku-daridda waqti-qaadaha qoraalka qoraalka
    • Using the model to identify a language
    • High-risk decision-making without appropriate verification