quiz bu lëkkale ak tegtal · Diggante Niveau

Laajteb dekodaasu jeton paralel MaskGIT

Saytu MaskGIT's maske-token taggat ak dekodage paralel iteratif, boole ci oraaru maskewaat bu sukkandiko ci wóolu boppam ak muy tekki ci ñaari yoon.

Yooni njiit yu jëm ciDekodage jeton paralel Maskgit
Laaj 1 de 8

naka la ko MaskGIT di dekode ay jeton nataal yu wuute ak transformatëru VQGAN?

Xoolaatal laaj bu nekk8 laaj

Sooy def quiz bi, tontu yi ak leeral yi dañuy feeñ.

  1. naka la ko MaskGIT di dekode ay jeton nataal yu wuute ak transformatëru VQGAN?
    • Dafay wax luy am ci token yi ci paralel ba noppi setal ci ay jéego yu néew
    • Dafay wax luy waaja am ci benn token ci yoon wu raster
    • Jëfandikuwul transformatër
    • Dafay defar mbind kese
  2. Ban mébetu tàggat la MaskGIT lebal ci misaali làkk yi?
    • Baat bi ci topp kese
    • Xalaatal token buñ maske, melni BERT
    • Xeetu nataal yi
    • Njàngale mu am doole ci feedback nit
  3. Ci diiru defar, ban token la MaskGIT di def ci iteration bu nekk?
    • Ñi gëna néew wóolu seen bopp
    • Jeton yu bari
    • Xalaatal yi gëna am wóolu seen bopp
    • Token yépp benn yoon ci jéego bu njëkk bi
  4. In the original MaskGIT generation examples, roughly how many refinement iterations were often used?
    • Exactly one iteration for every image and task
    • Hundreds of sequential token predictions
    • A fixed number that cannot be changed by the user
    • About 8 to 12 in those reported configurations
  5. Ban kalendriye mooy saytu ñaata token MaskGIT di wane ci jéego bu nekk?
    • Jamonoy maske cosinus
    • Regle bu lineer kese buy wane lépp benn yoon
    • Amul oraŋ; aleatër la
    • Benn jeton buñ tëral ci jéego bu nekk
  6. Lan moo waral bàyyi xel ci ñaari yoon jàppale MaskGIT?
    • Dafay nëbb nataal bi ci model bi
    • Dafay may token bu nekk mu gis nataal bi yépp ngir gëna am contexte
    • Dafay xool ci jeton yu njëkk ya
    • Dina dindi bépp soxla ci jeton
  7. What token grid does MaskGIT begin with when generating an image?
    • A fully decoded image that only needs color correction
    • A left-to-right stream with no masked positions
    • A grid with every image token masked
    • A grid of text embeddings from the prompt only
  8. How can MaskGIT fill a masked region during image inpainting?
    • Discard the observed image and generate an unrelated full image
    • Use only a left-to-right text decoder with no image tokens
    • Replace every visible token with a random codebook entry
    • Keep observed image tokens as context and iteratively predict the masked positions