Add tuple validation

app.py

@@ -17,6 +17,9 @@ language_model = PunctuationModel()
 
 def transcription(audio_stream, is_video = False):
     try:
+        if isinstance(audio_stream, tuple):
+            audio_stream = audio_stream[0]
+
         if is_video:
             waveform, sample_rate = torchaudio.load(audio_stream, format = "wav")
         else: