Bài kiểm tra có liên kết hướng dẫn · Trung bình Cấp độ

Câu đố DALL-E

Theo dõi các phiên bản lịch sử của DALL-E, với mã thông báo hình ảnh gốc, bộ giải mã và trước của DALL-E 2, chú thích, chỉnh sửa và ngừng sử dụng API của DALL-E 3.

Đường dẫn hướng dẫn liên quanDall E
Câu hỏi 1 của 8

What task is shared by the historical DALL-E family described here?

Xem trước mọi câu hỏi8 câu hỏi

Câu trả lời và giải thích xuất hiện khi bạn làm bài kiểm tra.

  1. What task is shared by the historical DALL-E family described here?
    • Text-to-image generation
    • Only speech translation
    • Only network-packet encryption
    • Only converting every photograph into a 3D mesh
  2. In the cited DALL-E 2 research, what does the text-conditioned prior produce for the decoder?
    • A mandatory raw audio waveform
    • Only an unchanged image filename
    • Every final output pixel directly without a decoder
    • CLIP image embeddings
  3. What did the historical DALL-E 3 ChatGPT integration do with a user’s brief image idea?
    • Retrained every original model weight during the request
    • Reduced every image to a single pixel
    • Refined it into a detailed generation prompt
    • Guaranteed that all instructions were satisfied perfectly
  4. Which training-data strategy is central to Improving Image Generation with Better Captions?
    • Deleting all text descriptions before training
    • Recaptioning training images with descriptive generated captions
    • Using only image filenames as every caption
    • Proving that every requested word renders perfectly
  5. What does inpainting mean in the historical image-editing context?
    • Only changing its file extension
    • Only adding an audio soundtrack
    • Automatically creating every possible camera view
    • Changing a selected region of an image
  6. Which generation design characterizes the original DALL-E research?
    • Autoregressive modeling of text and discrete image tokens
    • A required Gaussian-splat scene renderer
    • An unchanged lookup table of every future prompt
    • A model with no learned parameters
  7. What component provides the discrete image-token representation in the original DALL-E research?
    • A browser URL sorter
    • A discrete variational autoencoder
    • Only an automatic speech recognizer
    • Only a JSON grammar checker
  8. Which shutdown date does OpenAI’s API deprecation record list for both dall-e-2 and dall-e-3 snapshots?
    • May 12, 2025
    • Ngày 31 tháng 12 năm 2027
    • May 12, 2026
    • No removal date was listed