File size: 1,431 Bytes
eaffd42 |
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 |
from flask import Flask, request
import json
import base64
from faster_whisper import WhisperModel
import tempfile
import os
def base64_to_wav(base64_data, save_path):
wav_data = base64.b64decode(base64_data)
with open(save_path, 'wb') as file:
file.write(wav_data)
app = Flask(__name__)
GPU = True
wspr = WhisperModel("small", device="cuda" if GPU else "cpu", compute_type="int8")
@app.route('/api', methods=['GET' ,'POST'])
def STT():
if request.method == 'POST':
result = {}
audio_data = request.data
# Create a unique filename in the temporary directory
temp_dir = tempfile.gettempdir()
temp_file = tempfile.NamedTemporaryFile(suffix='.mp3', dir=temp_dir, delete=False)
save_path = temp_file.name
# save_path = "temp.wav"
# save_path = "temp.wav"
base64_to_wav(audio_data, save_path)
segments, info = wspr.transcribe(save_path, beam_size=5, language="fr")
texts = [el.text.strip() for el in segments]
transcription = texts[0]
query = transcription
result["transcription"] = query
if os.path.exists(save_path):
os.remove(save_path)
return result
elif request.method == 'GET':
# Add your test logic here
return "API is working correctly!"
if __name__ == '__main__':
app.run(host='0.0.0.0', port=5566, debug=True)
|