File size: 668 Bytes
57ea1c7 4504857 57ea1c7 4504857 57ea1c7 bc1c357 57ea1c7 |
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 |
import torch
from transformers import pipeline
model_id = 'distil-whisper/distil-large-v2'
pipe = pipeline(
"automatic-speech-recognition",
model=model_id,
chunk_length_s=15
)
# os.environ['PROTOCOL_BUFFERS_PYTHON_IMPLEMENTATION'] = 'python'
# os.environ['TRANSFORMERS_NO_ADVISORY_WARNINGS'] = '1'
# os.environ['TRANSFORMERS_VERBOSITY'] = 'error'
def score_audio(audio_path, true_result):
true_result = true_result.split('/')
transcription = pipe(audio_path)['text'].lower()
result = {'transcription': transcription,
'score': int(any([x.lower() in transcription for x in true_result])),
}
return result
|