mirror of
https://github.com/Monadical-SAS/reflector.git
synced 2026-03-31 19:36:47 +00:00
a682846645
* fix: local processing instead of http server for cpu * add fallback token if service worker doesnt work * chore: rename processors to keep processor pattern up to date and allow other processors to be createed and used with env vars
51 lines
1.5 KiB
Python
51 lines
1.5 KiB
Python
from faster_whisper import WhisperModel
|
|
|
|
from reflector.processors.audio_transcript import AudioTranscriptProcessor
|
|
from reflector.processors.audio_transcript_auto import AudioTranscriptAutoProcessor
|
|
from reflector.processors.types import AudioFile, Transcript, Word
|
|
from reflector.settings import settings
|
|
|
|
|
|
class AudioTranscriptWhisperProcessor(AudioTranscriptProcessor):
|
|
def __init__(self):
|
|
super().__init__()
|
|
self.model = WhisperModel(
|
|
settings.WHISPER_CHUNK_MODEL,
|
|
device="cpu",
|
|
compute_type="float32",
|
|
num_workers=12,
|
|
)
|
|
|
|
async def _transcript(self, data: AudioFile):
|
|
segments, _ = self.model.transcribe(
|
|
data.path.as_posix(),
|
|
language="en",
|
|
beam_size=5,
|
|
# condition_on_previous_text=True,
|
|
word_timestamps=True,
|
|
vad_filter=True,
|
|
vad_parameters={"min_silence_duration_ms": 500},
|
|
)
|
|
|
|
if not segments:
|
|
return
|
|
|
|
transcript = Transcript(words=[])
|
|
segments = list(segments)
|
|
ts = data.timestamp
|
|
|
|
for segment in segments:
|
|
for word in segment.words:
|
|
transcript.words.append(
|
|
Word(
|
|
text=word.word,
|
|
start=round(ts + word.start, 3),
|
|
end=round(ts + word.end, 3),
|
|
)
|
|
)
|
|
|
|
return transcript
|
|
|
|
|
|
AudioTranscriptAutoProcessor.register("whisper", AudioTranscriptWhisperProcessor)
|