Spaces:
Running
on
A10G
Running
on
A10G
File size: 1,753 Bytes
726ec90 2ec9baa 726ec90 6831f1f 726ec90 6831f1f 726ec90 9e4dc76 726ec90 2ec9baa 3d7f69e 726ec90 2ec9baa |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 |
import io
import threading
from multiprocessing import Queue
from queue import Empty
from faster_whisper import WhisperModel
import logging
import sys
# Configure logging
logging.basicConfig(
level=logging.INFO,
format="%(asctime)s - %(name)s - %(levelname)s - %(message)s",
handlers=[logging.StreamHandler(sys.stdout)],
)
# Get a logger for your app
logger = logging.getLogger(__name__)
class AudioTranscriber(threading.Thread):
def __init__(self, audio_queue: "Queue[io.BytesIO]", text_queue: "Queue[str]"):
super().__init__()
self.audio_queue = audio_queue
self.action_queue = text_queue
self.daemon = True # Thread will exit when main program exits
self.transcriber = WhisperModel(
"large",
device="cuda",
compute_type="int8",
)
def run(self):
while True:
try:
# Wait for 1 second before timing out and checking again
audio_chunk = self.audio_queue.get(timeout=1)
# Process the audio chunk using the faster-whisper implementation
segments, info = self.transcriber.transcribe(audio_chunk, language="fr")
# Put the transcription results in the output queue
for segment in segments:
self.action_queue.put(segment.text)
# Still print for debugging
logger.info(
f"[{segment.start:.2f}s -> {segment.end:.2f}s] {segment.text}"
)
except Empty:
continue # If queue is empty, continue waiting
except Exception as e:
logger.error(f"Error processing audio chunk: {e}")
|