quiz bu lëkkale ak tegtal · Diggante Niveau

Imagen Bind-ci-Nataal Quiz

Xoolaat Imagen bind bu njëkk bi, defar cascade, threshold dinaamik, ak tànneefi dëgër ci tàggat yaram bu am dayo bu mag.

Yooni njiit yu jëm ciImagen Bind ci Nataal
Laaj 1 de 8

In the original Imagen study's scaling comparison, which change produced larger image-quality and text-alignment gains?

Xoolaatal laaj bu nekk8 laaj

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

  1. In the original Imagen study's scaling comparison, which change produced larger image-quality and text-alignment gains?
    • Discarding the language encoder entirely
    • Increasing the pretrained text encoder's size rather than the diffusion model's size
    • Replacing diffusion with nearest-neighbor image lookup
    • Reducing every prompt to a single word
  2. Ban encoder text la Imagen di wéeru?
    • Reelu nataal buñ jaxasoo
    • LSTM bu ndaw
    • Regle yuñ bind ak loxo
    • ab xeetu làkk T5-XXL
  3. naka la ko Imagen mënee am dayo bu kawe?
    • Diffusion cascade ak etape yuy yokk dayo bu mag
    • Benn jéego bu mag
    • Dagg ab nataal bu yaatu
    • Ni ñuy boolee ñaari nataal
  4. What problem did original Imagen's dynamic thresholding address under strong guidance?
    • A lack of text tokens in the encoder vocabulary
    • The need to compress the final PNG file
    • Saturation of predicted pixel values during sampling
    • A missing training label for every object
  5. Luy banc dessin?
    • Tablet de desine
    • Màrku GPU
    • Done yu nataali muus
    • Benn référence buy wane ni ñuy jàngat xeetu bind-ci-nataal
  6. Lu tax ñuy tëye encodeur T5-XXL ci Imagen?
    • Dafa yàqu te mënul tàggat
    • Xam-xam bind bu am solo ci toxal tàggat yaram bu njëkk te amul beneen tuning
    • Ngir sakkanal bayaalu disk bi kese
    • Ndax bind bi amul benn njariñ
  7. What resolution did original Imagen's base diffusion model generate before its super-resolution stages?
    • 256 by 256 pixels as the first stage
    • 1024 by 1024 pixels directly
    • 64 by 64 pixels
    • A single pixel that was copied across the final image
  8. Why did the Imagen researchers add Gaussian noise to low-resolution conditioning images during super-resolution training?
    • To make the upsamplers more robust to imperfect inputs from earlier stages
    • To remove text conditioning from the entire cascade
    • To guarantee that all training images had the same caption
    • To replace the upsamplers with deterministic resizing