Wav2Vec 2.0

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

1. Ban jafe-jafe bu mag ci xàmmee kàddu lañu defar ngir saafara Wav2Vec 2.0?
   • Audio buñ etikete (ñu bind)
   • Fichier audio yi dañu yaatu lool ngir denc leen
   • Mikro yi seer nañu lool
   • Text bi jafe tokenise
2. Ban xeetu mébetu jàng la Wav2Vec 2.0 di jëfandikoo ci diiru tàggat-yaram bi?
   • Xeetu fonem yu jub ak etiket yu mat
   • ab ñakk bu wuute ci kaw segment yu nëbbu yuñ maske
   • Tabaxaat pixel
   • Njàngale mu am doole ci feedback nit
3. Ban lëf mooy njëkka defar jëmmu onde bu ñor ci Wav2Vec 2.0?
   • LSTM buy dellusi
   • ab échantillonnage diffusion
   • Encodeur man-manu CNN bu bari ay layer
   • ab extracteur MFCC buñu defaree loxo
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. Lan mooy wareefu téere kode biñ jàng ci Wav2Vec 2.0?
   • Dafay xaaj audio bu wéy ci kàddu yu am àpp
   • Dafay denc baatu-jàll jëfandikukat yi
   • Dafay tënk transkripsioŋ bu mujj bi
   • Dafay tànn njariñu mikro bi
6. Ban xeetu làkk yu bari lañu tabax ci kaw Wav2Vec 2.0?
   • 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