Expands the note for distil-whisper (#659)

This commit is contained in:
Purfview
2024-01-28 20:48:40 +00:00
committed by GitHub
parent 3aec421849
commit f144e4c83d

View File

@@ -170,7 +170,7 @@ segments, info = model.transcribe("audio.mp3", beam_size=5,
language="en", max_new_tokens=128, condition_on_previous_text=False)
```
NOTE: emprically, `condition_on_previous_text=True` will degrade the performance of `faster-distil-whisper` for long audio.
NOTE: emprically, `condition_on_previous_text=True` will degrade the performance of `faster-distil-whisper` for long audio. Degradation on the first chunk was observed with `initial_prompt` too.
### Word-level timestamps