Skip to content
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

Jump to bottom

Add support for Faster Whisper #693

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

Add support for Faster Whisper #693

Changes from 1 commit
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
1ec6b45
Add support for Faster Whisper
joshuaboniface Jun 5, 2023
b558ab6
Add tests identical to normal whisper
joshuaboniface Jun 5, 2023
1bd5e48
Correct tests for faster_whisper
joshuaboniface Jun 5, 2023
83b6905
Allow setting download root cache path
joshuaboniface Jun 6, 2023
f0c7fb1
Move faster_whisper transcribing to subprocess
joshuaboniface Jun 9, 2023
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
Prev Previous commit
Next Next commit
Allow setting download root cache path
  • Loading branch information
@joshuaboniface
joshuaboniface committed Jun 6, 2023
commit 83b6905d16fc84abfa00fd4f2d17be7e552242a5
6 changes: 4 additions & 2 deletions speech_recognition/recognizers/fasterwhisper.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -11,6 +11,7 @@ def recognize_faster_whisper(
recognizer,
audio_data: "AudioData",
model="base",
download_root=None,
beam_size=5,
device="auto",
compute_type="int8",
Expand All @@ -29,6 +30,7 @@ def recognize_faster_whisper(
{beam_size}: Allows overriding the default beam size if desired.
{device}: Allows setting the device, e.g. "auto", "cpu", "cuda", etc.
{compute_type}: Allows setting the compute type, e.g. "int8", "float16", etc.
{download_root}: Allows setting the model cache root path

The return dict is also customized to avoid returning a mess of FasterWhisper objects.
"""
Expand All @@ -39,9 +41,9 @@ def recognize_faster_whisper(
import soundfile as sf
import faster_whisper

print(f"model: {model}, beam_size: {beam_size}, device: {device}, compute_type: {compute_type}, language: {language}, translate: {translate}")
print(f"model: {model}, download_root: {download_root}, beam_size: {beam_size}, device: {device}, compute_type: {compute_type}, language: {language}, translate: {translate}")

whisper_model = faster_whisper.WhisperModel(model, device=device, compute_type=compute_type)
whisper_model = faster_whisper.WhisperModel(model, device=device, compute_type=compute_type, download_root=download_root)

# 16 kHz https://github.com/openai/whisper/blob/28769fcfe50755a817ab922a7bc83483159600a9/whisper/audio.py#L98-L99
wav_bytes = audio_data.get_wav_data(convert_rate=16000)
Expand Down