Quiz collegato alla guida · Difficile Livello

Quiz sussurro per la messa a punto

Pianifica l'adattamento di Whisper utilizzando coppie audio-testo allineate, suddivisioni pulite degli altoparlanti, compromessi in termini di dimensioni del modello e una solida valutazione per dimenticare.

Percorsi guida correlatiSussurro di sintonia fine
Domanda 1 di 8

Cosa cambia durante la messa a punto supervisionata di Whisper?

Anteprima di ogni domanda8 domande

Le risposte e le spiegazioni vengono visualizzate mentre rispondi al quiz.

  1. Cosa cambia durante la messa a punto supervisionata di Whisper?
    • Only the decoded transcript is edited after inference
    • Only the language label is changed without training
    • Model weights are updated using paired audio and text examples
    • Only a static list of spellings is appended to the output
  2. Perché i segmenti audio e trascrizione devono essere allineati?
    • Alignment guarantees the same GPU memory use for every segment
    • Misaligned text teaches the model incorrect speech-to-text mappings
    • An alignment error is harmless if the transcript contains domain vocabulary
    • Correct timing guarantees that every transcript word is accurate
  3. Per valutare le prestazioni su interlocutori invisibili quando ciascun oratore contribuisce con molte registrazioni, come dovrebbero essere partizionati i dati?
    • Randomly split overlapping windows from each speaker
    • Place every speaker in every partition
    • Use test speakers during training only
    • Group by speaker before creating train, validation and test splits
  4. Cosa supporta un set di validazione sospeso durante la messa a punto?
    • Checkpoint and hyperparameter decisions
    • Unlimited tuning followed by an unbiased final score on that same set
    • Gradient updates using every held-out label
    • A guarantee that the final deployment distribution will be unchanged
  5. Cosa può differire tra i rapporti sul tasso di errore di due parole?
    • Only the number of model parameters
    • Only the loss observed on the training set
    • Text normalization and tokenization conventions
    • Only whether both systems use the same tokenizer name
  6. Why monitor a general-domain set while adapting to a narrow domain?
    • General data always lowers target accuracy
    • Target adaptation can reduce performance outside its training distribution
    • The test set should be used for gradient updates
    • The model size changes automatically
  7. What can low-rank adapters offer when the selected implementation supports them?
    • Guaranteed lower WER in every domain
    • Automatic correction of mislabeled transcripts
    • A guarantee that no general-domain behavior can change
    • Fewer trainable parameters during adaptation
  8. Why compare fine-tuning with decoding adjustments or domain-term support?
    • Training may add cost when simpler changes solve the problem
    • Decoding choices always change model weights
    • Fine-tuning cannot affect vocabulary behavior
    • A baseline makes test labels unnecessary