Update app.py

app.py

@@ -9,6 +9,7 @@ import torchaudio
 from transformers import AutoConfig, AutoModel, Wav2Vec2FeatureExtractor
 import librosa
 import numpy as np
+import subprocess
 
 
 def speech_file_to_array_fn(path, sampling_rate):
@@ -42,7 +43,9 @@ model.to(device)
 
 
 def transcribe(audio):
-    return predict(audio, 16000)
+    command = f"sox -t wav {audio} -r 16000 -b 16 c_{audio} channels 1"
+    subprocess.call(command, shell=False)
+    return predict(f"c_{audio}", 16000)
 
 
 def get_asr_interface():