Issues

• Replicate experiments on wave2vec

  AIRC-ASR/partial-asr#37 · Filter by author omarASC5 opened on Oct 7, 2023

• Copy over the relevant code from the GrammarGPT repo to partial-asr repo

  AIRC-ASR/partial-asr#36 · Filter by author omarASC5 opened on Oct 7, 2023

• Evaluate 1000 data points, 5 seconds removed, model on all LibriSpeech test and dev sets

  AIRC-ASR/partial-asr#35 · Filter by author omarASC5 opened on Oct 7, 2023

• Evaluate 1000 data points, 3 second removed, model on all LibriSpeech test and dev sets

  AIRC-ASR/partial-asr#34 · Filter by author omarASC5 opened on Oct 7, 2023

• Evaluate 1000 data points, 4 seconds removed, model on all LibriSpeech test and dev sets

  AIRC-ASR/partial-asr#33 · Filter by author omarASC5 opened on Oct 7, 2023

• Evaluate 1000 data points, 2 seconds removed, model on all LibriSpeech test and dev sets

  AIRC-ASR/partial-asr#32 · Filter by author omarASC5 opened on Oct 7, 2023

• Evaluate 1000 data points, 1 second removed, model on all LibriSpeech test and dev sets

  AIRC-ASR/partial-asr#31 · Filter by author omarASC5 opened on Oct 7, 2023

• Evaluate 1000 data points, no audio removed, model on all LibriSpeech test and dev sets

  AIRC-ASR/partial-asr#30 · Filter by author omarASC5 opened on Oct 7, 2023

• Train model on full LibriSpeech, 4 epochs, 5 seconds removed

  AIRC-ASR/partial-asr#29 · Filter by author omarASC5 opened on Oct 7, 2023

• Train model on full LibriSpeech, 4 epochs, 4 seconds removed

  AIRC-ASR/partial-asr#28 · Filter by author omarASC5 opened on Oct 7, 2023

• Train model on full LibriSpeech, 4 epochs, 3 seconds removed

  AIRC-ASR/partial-asr#27 · Filter by author omarASC5 opened on Oct 7, 2023

• Train model on full LibriSpeech, 4 epochs, 2 seconds removed

  AIRC-ASR/partial-asr#26 · Filter by author omarASC5 opened on Oct 7, 2023

• Train model on full LibriSpeech, 4 epochs, 1 second removed

  AIRC-ASR/partial-asr#25 · Filter by author omarASC5 opened on Oct 7, 2023

• Train model on full LibriSpeech, 4 epochs, no audio removed

  AIRC-ASR/partial-asr#24 · Filter by author omarASC5 opened on Oct 7, 2023

• Train model on 1000 data points, 15 epochs, 5 seconds removed

  AIRC-ASR/partial-asr#23 · Filter by author omarASC5 opened on Oct 7, 2023

• Train model on 1000 data points, 15 epochs, 4 seconds removed

  AIRC-ASR/partial-asr#22 · Filter by author omarASC5 opened on Oct 7, 2023

• Train model on 1000 data points, 15 epochs, 3 seconds removed

  AIRC-ASR/partial-asr#21 · Filter by author omarASC5 opened on Oct 7, 2023

• Train model on 1000 data points, 15 epochs, 2 seconds removed

  AIRC-ASR/partial-asr#20 · Filter by author omarASC5 opened on Oct 7, 2023

• Train model on 1000 data points, 15 epochs, 1 second removed

  AIRC-ASR/partial-asr#19 · Filter by author omarASC5 opened on Oct 7, 2023

• Train model on 1000 data points, 15 epochs, no audio removed

  AIRC-ASR/partial-asr#18 · Filter by author omarASC5 opened on Oct 7, 2023

• Make a prompt for correcting partial audio files using ASR N-best lists (N=5)

  AIRC-ASR/partial-asr#17 · Filter by author omarASC5 opened on Oct 7, 2023

• Run our untrained model 0-shot on 5 seconds removed from middle 10 seconds

  AIRC-ASR/partial-asr#16 · Filter by author omarASC5 opened on Oct 7, 2023

• Run ChatGPT 0-shot on 5 seconds removed from middle 10 seconds

  AIRC-ASR/partial-asr#15 · Filter by author omarASC5 opened on Oct 7, 2023

• Run our untrained model 0-shot on 4 seconds removed from middle 10 seconds

  AIRC-ASR/partial-asr#14 · Filter by author omarASC5 opened on Oct 7, 2023

• Run ChatGPT 0-shot on 4 seconds removed from middle 10 seconds

  AIRC-ASR/partial-asr#13 · Filter by author omarASC5 opened on Oct 7, 2023