Update app.py

app.py

@@ -39,26 +39,26 @@ def plot_sound_event_detection_result(framewise_output):
 def pred(audio):
     rate, y = audio
     device = 'cpu' # 'cuda' | 'cpu'
-    print('sample rate ', rate)
-    print('shape ', y.shape)
-    print('raw data', y)
+    #print('sample rate ', rate)
+    #print('shape ', y.shape)
+    #print('raw data', y)
     y = y.astype(np.float32)
-    print('float', y)
+    #print('float', y)
     y = librosa.core.to_mono(y.T)
-    print('shape ', y.shape)
-    print('mono', y)
+    #print('shape ', y.shape)
+    #print('mono', y)
     y = librosa.core.resample(y, orig_sr=rate, target_sr=32000)
-    print('shape ', y.shape)
-    print('resampled', y)
-    print(y.mean())
-    print(y.std())
+    #print('shape ', y.shape)
+    #print('resampled', y)
+    #print(y.mean())
+    #print(y.std())
     #y = (y - y.mean())/y.std()
     y = y/y.max()
-    print('normalized', y)
+    #print('normalized', y)
     #print(rate)
-    plt.plot(y)
-    plt.savefig('wave.png')
-    plt.close()
+    #plt.plot(y)
+    #plt.savefig('wave.png')
+    #plt.close()
     y = y[None, :]  # (batch_size, segment_samples)
     #print(y)
     
@@ -86,7 +86,6 @@ demo = gr.Interface(
         "telephone_speech.wav",
         "ringtone.wav", "animals.wav",
     ],
-    interpretation="default",
 )
 
 demo.launch()