quiz bu lëkkale ak tegtal · Diggante Niveau

Diffusion Personnalisé Multi-Concept Quiz

Xoolaat tuning cross-attention bu Custom Diffusion, personalisasioŋ bu néew, yamale, ak anam yu ñuy jàppale ngir bind konsept yu bari.

Yooni njiit yu jëm ciDiffusion Personnalisé Multi Concept Tuning
Laaj 1 de 8

Ban wàll ci xeetu diffusion bi Diffusion Personnalisé di njëkka defar?

Xoolaatal laaj bu nekk8 laaj

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

  1. Ban wàll ci xeetu diffusion bi Diffusion Personnalisé di njëkka defar?
    • Caabi ak projection valeur ci diggu bàyyi xel
    • Bépp diisaay ci U-Net
    • Decodeur VAE bi kese
    • Tokenisatëru bind bi
  2. Lan mooy Diffusion Personnalisée bi gëna fës buñu ko méngale ak pexe yu am benn konsept?
    • Defar wideo ci palaasu nataal
    • Defar ay konsept yu bees yu bari ci benn seen
    • Soxlawul benn nataal
    • Dafay daw ci CPU yi kese
  3. What minimum number of training images did the Custom Diffusion paper report using in some experiments?
    • As few as four images
    • At least 500 images
    • Exactly one image in every experiment
    • The paper specifies no concept images
  4. Lan moo waral Custom Diffusion di jëfandikoo ay nataali regularisasioŋ yu bawoo ci kategori bu gëna yaatu ci konsept bi?
    • ngir gaawlu VAE
    • Ngir yokk dayo nataal bi
    • Ngir dindi filigran yi
    • Ngir moytu overfitting ak làkk drift ci baat konsept
  5. naka lañu mëna boolee ñaari konsepti Custom Diffusion yuñ tàggatee ci anam wu wuute?
    • Buñu leen boole jaaraleko ci ab optimisation buñu tënk ci kaw poid yiñ akorde
    • Mënu leen boole
    • ci moyenne nataal yiñ dugal
    • Ni ñuy tàggataat model bi yépp ci noonu
  6. What proportion of model weights did the Custom Diffusion paper report storing for its tuned subset?
    • About 3% of the model weights
    • All model weights plus a new model copy
    • No trained parameters at all
    • A fixed 50% of the model weights
  7. Besides merging separately tuned models, how does Custom Diffusion support learning multiple concepts?
    • It requires a separate foundation model for every prompt
    • It replaces all cross-attention with a fixed lookup table
    • It can train for multiple concepts jointly
    • It can only learn one concept per model
  8. In a cross-attention layer, what do the tuned text-side key and value projections help associate?
    • Image pixels with internet search results
    • The diffusion timestep with the model's source code
    • Audio frequencies with a document's page number
    • Prompt concepts with features in the image latent representation