Guide-yakabatana mibvunzo · Pakati Nhanho

Mudzidzisi Kumanikidza Mukutevedzana Mamodeli Mibvunzo

Inotarisa kuti zvirevo zvepasi-chokwadi zvinodzidzisa sei decoder inoteedzana, masikirwo ezvikonzero zvinotsigira kudzidziswa kwakafanana, uye nei kufungidzira kuchisiyana.

Nzira dzekutungamira dzinoenderanaMudzidzisi Kumanikidza Mukutevedzana Mamodheru
Mubvunzo 1 ye 8

Panguva yekumanikidza mudzidzisi, chii chinodyiswa sechipimo padanho rega rega rekudhikodha?

Tarisa mubvunzo wese8 mibvunzo

Mhinduro uye tsananguro zvinoonekwa paunenge uchitora bvunzo.

  1. Panguva yekumanikidza mudzidzisi, chii chinodyiswa sechipimo padanho rega rega rekudhikodha?
    • Ivhu-chokwadi chekare chiratidzo
    • Iyo modhi yakare kufanotaura
    • Chiratidzo chisina kujairika
    • Chiratidzo chekupedzisira chekubuda
  2. During teacher-forced training, what target is paired with decoder input y_(t−1)?
    • The previous ground-truth token
    • The next target token y_t
    • A token sampled from the validation set
    • The maximum-probability output from an earlier training run
  3. Kusawirirana kwechitima-bvunzo kunokonzerwa nekumanikidza kwemudzidzisi kunozivikanwa se:
    • Vanishing gradients
    • Mode kudonha
    • Exposure bias
    • Label leakage
  4. MuTransformer decoder, ndeipi nzira inoita kuti kudzidziswa kwekumanikidzwa kwevadzidzisi kumhanye kwakafanana munzvimbo dzese?
    • Dropout
    • Batch normalization
    • Beam search
    • A causal (tarisa-mberi) yekutarisisa mask
  5. Panguva yekufungidzira, nei kumanikidza kwemudzidzisi kusingagone kushandiswa?
    • Iko hakuna pasi-chokwadi remangwana zviratidzo zvekudyisa
    • Basa rekurasikirwa harisipo
    • Iyo GPU inononoka
    • Masks ekutarisisa akaremara
  6. Ndeipi kurasikirwa kunowanzo shandiswa padanho rega rega panguva yekudzidziswa-kumanikidzwa mudzidzisi?
    • Kureva kukanganisa kwakapetwa kakapetwa kaviri pamarogi mbishi
    • Muchinjika-entropy pakati pezvakafanotaurwa kugovera uye chechokwadi chinotevera tokeni
    • Hinge kurasikirwa pane embeddings
    • Cosine kufanana kurasikirwa
  7. The target sequence is “red, blue, green.” Under teacher forcing, which token is supplied when training the prediction of “green”?
    • The model’s prediction for “green”
    • The future token after “green”
    • The ground-truth previous token “blue”
    • A randomly sampled vocabulary item
  8. At decoder position t, which tokens must a causal attention mask block?
    • The source sequence
    • All earlier target tokens
    • The current position’s own input
    • Target tokens at positions later than t