Guide-yakabatana mibvunzo · Pakati Nhanho

Wav2Vec 2.0 Quiz

Edzai Wav2Vec 2.0's inozvitarisisa pretraining resipi, inosiyanisa chinangwa, uye miganho yeyakaderera-label LibriSpeech mhedzisiro.

Nzira dzekutungamira dzinoenderanaWav2vec 2 0
Mubvunzo 1 ye 8

Nderipi dambudziko guru mukuzivikanwa kwekutaura raive Wav2Vec 2.0 rakagadzirirwa kugadzirisa?

Tarisa mubvunzo wese8 mibvunzo

Mhinduro uye tsananguro zvinoonekwa paunenge uchitora bvunzo.

  1. Nderipi dambudziko guru mukuzivikanwa kwekutaura raive Wav2Vec 2.0 rakagadzirirwa kugadzirisa?
    • Iko kushomeka kwemaodhiyo akanyorwa (akanyorwa).
    • Mafaira ekurekodhwa arikurisa kuchengetwa
    • Maikorofoni ari kudhura zvakanyanya
    • Mameseji ari kunetsa kuisa tokeni
  2. Ndeupi rudzi rwechinangwa chekudzidza chinoshandiswa neWav2Vec 2.0 panguva yekudzidzira?
    • Yakananga phoneme classification ine mavara akazara
    • Kurasikirwa kwakasiyana pane masked latent segment
    • Pixel kuvakazve
    • Kusimbisa kudzidza kubva kumhinduro dzevanhu
  3. Ndechipi chikamu chekutanga chinogadzirisa iyo yakaomeswa waveform muWav2Vec 2.0?
    • Idzokororo LSTM
    • A diffusion sampler
    • Iyo yakawanda-layer CNN inoratidzira encoder
    • MFCC inodhonza nemaoko
  4. In a reported LibriSpeech experiment, what small labeled-data setup demonstrated low-label potential?
    • 10 minutes unlabeled with no pretraining
    • 53,000 hours labeled with no pretraining
    • One hour labeled with no unlabeled data
    • 10 minutes labeled after pretraining on 53,000 hours of unlabeled speech
  5. Nderipi basa rebhuku rekodhi rakadzidzwa muWav2Vec 2.0?
    • Iyo inosiyanisa inoenderera odhiyo kuita inogumira kutaura zvikamu
    • Inochengeta mapassword evashandisi
    • Inomanikidza chinyorwa chekupedzisira
    • Inosarudza maikorofoni kuwana
  6. Ndeipi yakakura modhi yemitauro yakawanda yakavakirwa paWav2Vec 2.0 maitiro?
    • GPT-2
    • XLS-R
    • Stable Diffusion
    • ResNet-50
  7. What does the Transformer contribute during Wav2Vec 2.0 pretraining?
    • Converts transcripts into audio
    • Clusters final text into language labels
    • Contextualizes latent speech features around masked spans
    • Removes the feature encoder
  8. What is typically done after self-supervised pretraining to create an ASR model?
    • Replace the audio encoder with an image classifier
    • Train only on unlabeled text
    • Discard the model and use k-means alone
    • Fine-tune with transcribed speech