Skip to content

Bugfix: Fix inconsistent VAD defaults in BatchedInferencePipeline #1409

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
Purfview wants to merge 5 commits into
base: master
Choose a base branch
from
Open

Bugfix: Fix inconsistent VAD defaults in BatchedInferencePipeline #1409

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
d5dda8d
Bugfix: Fix inconsistent VAD defaults in BatchedInferencePipeline
Purfview Dec 31, 2025
8d58094
Fix
Purfview Dec 31, 2025
e8f1ac2
Add comment to vad.py
Purfview Dec 31, 2025
d80c6f3
Whitespace fix
Purfview Dec 31, 2025
831c128
Fix 2
Purfview Dec 31, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
18 changes: 8 additions & 10 deletions faster_whisper/transcribe.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -395,18 +395,16 @@ def transcribe(
# if no segment split is provided, use vad_model and generate segments
if not clip_timestamps:
if vad_filter:
batched_defaults = VadOptions(
max_speech_duration_s=chunk_length,
min_silence_duration_ms=160,
)
if vad_parameters is None:
vad_parameters = VadOptions(
max_speech_duration_s=chunk_length,
min_silence_duration_ms=160,
)
vad_parameters = batched_defaults
elif isinstance(vad_parameters, dict):
if "max_speech_duration_s" in vad_parameters.keys():
vad_parameters.pop("max_speech_duration_s")

vad_parameters = VadOptions(
**vad_parameters, max_speech_duration_s=chunk_length
)
merged_vad_params = {**asdict(batched_defaults), **vad_parameters}
merged_vad_params["max_speech_duration_s"] = chunk_length
vad_parameters = VadOptions(**merged_vad_params)

clip_timestamps = get_speech_timestamps(audio, vad_parameters)
# run the audio if it is less than 30 sec even without clip_timestamps
Expand Down
2 changes: 1 addition & 1 deletion faster_whisper/vad.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -42,7 +42,7 @@ class VadOptions:
neg_threshold: float = None
min_speech_duration_ms: int = 0
max_speech_duration_s: float = float("inf")
min_silence_duration_ms: int = 2000
min_silence_duration_ms: int = 2000 # defaults to 160 in BatchedInferencePipeline
speech_pad_ms: int = 400
min_silence_at_max_speech: int = 98
use_max_poss_sil_at_max_speech: bool = True
Expand Down