Skip to content

Add duration of audio and VAD removed duration to BatchedInferencePipeline #1186

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

Merged
MahmoudAshraf97 merged 14 commits into from
Dec 23, 2024
Merged

Add duration of audio and VAD removed duration to BatchedInferencePipeline #1186

Changes from all commits
Commits
Show all changes
14 commits
Select commit Hold shift + click to select a range
3626f6a
initial commit
MahmoudAshraf97 Oct 31, 2024
5f905ee
remove torch from reqirements
MahmoudAshraf97 Oct 31, 2024
65dc596
fix formatting and dtype
MahmoudAshraf97 Oct 31, 2024
dcc95af
add type annotations
MahmoudAshraf97 Oct 31, 2024
f6adb22
remove CuPy
MahmoudAshraf97 Oct 31, 2024
bc86503
reduce padding to `hop_length` insteadn of `n_samples`
MahmoudAshraf97 Oct 31, 2024
9f4124c
Merge remote-tracking branch 'MahmoudAshraf97/numpy_fe'
greenw0lf Nov 6, 2024
fdadc64
Attempt to fix logging for batched Whisper
greenw0lf Nov 6, 2024
15c39f4
Hopefully same logging as with normal WhisperModel
greenw0lf Nov 7, 2024
15add92
Merge remote-tracking branch 'MahmoudAshraf97/master'
greenw0lf Dec 3, 2024
0a119b4
Merge branch 'SYSTRAN:master' into master
greenw0lf Dec 3, 2024
ef8c2ce
Revert irrelevant changes
greenw0lf Dec 3, 2024
24448f6
Avoid duplicate logger
greenw0lf Dec 23, 2024
ecd74c3
Replace self.logger with self.model.logger
greenw0lf Dec 23, 2024
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
9 changes: 9 additions & 0 deletions faster_whisper/transcribe.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -388,6 +388,10 @@ def transcribe(
audio = decode_audio(audio, sampling_rate=sampling_rate)
duration = audio.shape[0] / sampling_rate

self.model.logger.info(
"Processing audio with duration %s", format_timestamp(duration)
)

chunk_length = chunk_length or self.model.feature_extractor.chunk_length
# if no segment split is provided, use vad_model and generate segments
if not clip_timestamps:
Expand Down Expand Up @@ -421,6 +425,11 @@ def transcribe(
/ sampling_rate
)

self.model.logger.info(
"VAD filter removed %s of audio",
format_timestamp(duration - duration_after_vad),
)

audio_chunks, chunks_metadata = collect_chunks(audio, clip_timestamps)
features = (
[self.model.feature_extractor(chunk)[..., :-1] for chunk in audio_chunks]
Expand Down
Loading