Update app.py

app.py

@@ -13,24 +13,17 @@ for m in models.values():
     m.model.generation_config.input_ids = m.model.generation_config.forced_decoder_ids
     m.model.generation_config.forced_decoder_ids = None
 
-
 def ensure_mono_16k(audio_path):
-    """Load audio, convert to mono + 16kHz, and save a temp version"""
+    """Convert audio to mono + 16 kHz"""
     waveform, sr = torchaudio.load(audio_path)
-    
-    # Convert to mono if necessary
     if waveform.shape[0] > 1:
         waveform = waveform.mean(dim=0, keepdim=True)
-    
-    # Resample to 16kHz if necessary
     if sr != 16000:
         resampler = torchaudio.transforms.Resample(sr, 16000)
         waveform = resampler(waveform)
         sr = 16000
-    
-    tmp_path = "/tmp/processed_16k.wav"
-    torchaudio.save(tmp_path, waveform, sr)
-    return tmp_path
+    return waveform, sr
+
 
 def trim_leading_silence(waveform, sr, keep_ms=100, threshold=0.01):
     """