Idanwo ti o sopọ mọ itọsọna · Alabọde Ipele

Encoder-Decoder Architectures adanwo

Review encoder-decoder model components, source-to-target attention and text-to-text uses in translation and language tasks.

Jẹmọ itọsọna awọn ọnaEncoder Decoder Architectures
Ibeere 1 ti 8

Kini iṣẹ akọkọ ti kooduopo ni awoṣe koodu-iyipada kan?

Awotẹlẹ gbogbo ibeere8 ibeere

Awọn idahun ati awọn alaye han bi o ṣe n mu ibeere naa.

  1. Kini iṣẹ akọkọ ti kooduopo ni awoṣe koodu-iyipada kan?
    • Lati ka igbewọle ati yi pada si awọn aṣoju inu ọlọrọ
    • Lati ṣe agbejade iṣẹjade kan aami ni akoko kan
    • Lati ipo awọn abajade oludije nipasẹ iṣeeṣe
    • Lati tọju data ikẹkọ awoṣe
  2. Kilode ti olupilẹṣẹ ṣe lo boju-boju (okunfa) akiyesi ara ẹni?
    • Lati tọju titẹ sii lati koodu koodu
    • Lati ṣe idiwọ ipo kọọkan lati rii awọn ami ami iwaju lakoko iran
    • Lati compress awọn ti o wu sinu díẹ àmi
    • Lati fi awọn ami silẹ laileto fun isọdọtun
  3. Ilana wo ni o so kooduopo pọ mọ awọn aṣoju kooduopo?
    • Jade kuro
    • Ipese deede
    • Cross-akiyesi
    • Ifaminsi ipo
  4. Eyi ti awọn awoṣe jẹ ẹya kooduopo-decoder (akọọkan-si-ọkọọkan) faaji?
    • GPT-3
    • BERT
    • Ọrọ2Vec
    • T5
  5. Kini idi ti apẹrẹ koodu-iyipada jẹ ibamu adayeba fun itumọ?
    • Nitori igbewọle ati iṣejade yatọ si awọn ọna ṣiṣe, to nilo oye ni kikun lẹhinna iran tuntun
    • Nitori titẹ sii ati iṣelọpọ nigbagbogbo jẹ ipari kanna
    • Nitoripe ko nilo lati ṣe ipilẹṣẹ ọrọ tuntun rara
    • Nitoripe o yago fun lilo akiyesi patapata
  6. In the original Transformer architecture, what attention does the encoder use over its input sequence?
    • Causal attention that hides all earlier source tokens
    • Bidirectional self-attention over the full source sequence
    • Cross-attention only to an image sequence
    • No attention mechanism
  7. What training objective does BART use for pretraining?
    • It predicts only the next pixel in an image
    • It classifies each word into a fixed sentiment label
    • It corrupts text and trains a sequence-to-sequence denoising model to reconstruct the original
    • It performs speech recognition from raw audio
  8. How does T5 frame many NLP tasks for one encoder-decoder model?
    • As image classification with a shared label set
    • As speech waveforms with no text tokens
    • As separate model architectures for each task
    • As text-to-text inputs and outputs with task descriptions in the input text