Quiz Space

Deep Learning Practice · Quiz 2 · 16 Aug 2026 · May 2026 term

Question 15: Why might you freeze the encoder layers of a pre-trained…

Question 15

+3 marksOne correct option

Why might you freeze the encoder layers of a pre-trained ASR model during the initial phases of fine-tuning on a very small dataset ?

  1. A

    Reducing the inference latency.

  2. B

    Eliminating the need for CTC.

  3. C

    Preventing catastrophic forgetting.

  4. D

    Increasing the model capacity.

Show answer

Correct answer

  • C

    Preventing catastrophic forgetting.

See the whole paperTake it as a mock test

Question 15 of 19 in the IIT Madras BS Deep Learning Practice (Deep Learning Practice) Quiz 2 paper sat on 16 Aug 2026, in the May 2026 term (Deep Learning Practice 16 Aug 26). It carries 3 marks.

More questions from this paper

  1. Q1Which of the following factors can adversely impact the accuracy of speech language identification?
  2. Q2You are building an end-to-end speaker-attributed transcription pipeline using the Whisper model for speech recognition…
  3. Q3When feeding audio into the pretrained speechbrain/spkrec-ecapa-voxceleb model, which of the following statements regar…
  4. Q4Which statements accurately describe the roles, constraints, and data-processing behaviors when preparing batches for s…
  5. Q5An audio recording is sampled at 16,000 Hz for 8 seconds. How many samples are contained in the recording?
  6. Q6A multilingual language identification model is trained to classify 40 languages. The training dataset contains 1000 au…
  7. Q7An Automatic Speech Recognition (ASR) system produces a hypothesis transcript for a spoken audio clip. Calculate the Wo…
  8. Q8In a spoken language identification task, a pretrained speech model is used to extract features from an input audio wav…
  9. Q9What is the effect of setting compute_type="int8" in WhisperModel(whisper_model, compute_type="int8")?
  10. Q10Consider the following incomplete script intended for fine-tuning a Wav2Vec2 model: Which of the following code blocks …
  11. Q11Which of the following describes the correct input for the Decoder during the training of a Seq2Seq ASR model like Whis…
  12. Q12When fine-tuning a pre-trained Transformer-based ASR model using Connectionist Temporal Classification (CTC), what is t…
  13. Q13The input tensor train_data has shape Determine the correct value of in_features for the fc1 layer.
  14. Q14Figure question
  15. Q16What is the primary disadvantage of using a character-level tokenizer compared to a Byte-Pair Encoding (BPE) subword to…
  16. Q17In a standard TTS pipeline involving an acoustic model and a vocoder, which component is primarily responsible for conv…
  17. Q18Which of the following linkages is possible in agglomerative clustering?
  18. Q19The "sentence-transformers/all-mpnet-base-v2" model generates 768-dimensional sentence embeddings. A data processing pi…