You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
LibriVox recordings usually begin with a stock introduction such as the following one for English:
"This is a LibriVox recording. All LibriVox recordings are in the public domain. For more information or to volunteer, visit librivox.org."
It seems that when asked to transcribe a LibriVox recording, Whisper will tend to skip this stock intro and transcribe only the part for which the transcript was available in the LibriVox metadata.
We could make some guesses about what the training data might look like to cause this, although here I just wanted to note the observation.
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
Uh oh!
There was an error while loading. Please reload this page.
-
LibriVox recordings usually begin with a stock introduction such as the following one for English:
"This is a LibriVox recording. All LibriVox recordings are in the public domain. For more information or to volunteer, visit librivox.org."
It seems that when asked to transcribe a LibriVox recording, Whisper will tend to skip this stock intro and transcribe only the part for which the transcript was available in the LibriVox metadata.
We could make some guesses about what the training data might look like to cause this, although here I just wanted to note the observation.
Beta Was this translation helpful? Give feedback.
All reactions