Update app.py

app.py

@@ -48,20 +48,17 @@ def generate_segment_audio(text, lang, speaker_url, pipe):
     return audio_np
 
 def concatenate_audio_segments(segments):
-    stereo_segments = []
-    for seg in segments:
-        if seg.ndim == 1 or seg.shape[1] == 1:
-            stereo_seg = np.stack((seg, seg), axis=-1)
+    concatenated_audio_data = []
+    for segment in segments:
+        if segment.ndim == 1:
+            stereo_segment = np.stack((segment, segment), axis=-1)
+        elif segment.shape[1] == 1:
+            stereo_segment = np.concatenate((segment, segment), axis=1)
         else:
-            stereo_seg = seg
-        stereo_segments.append(stereo_seg)
-    total_length = sum(seg.shape[0] for seg in stereo_segments)
-    concatenated_audio = np.zeros((total_length, 2), dtype=np.float32)
-    current_index = 0
-    for seg in stereo_segments:
-        end_index = current_index + seg.shape[0]
-        concatenated_audio[current_index:end_index, :] = seg
-        current_index = end_index
+            stereo_segment = segment
+
+        concatenated_audio_data.append(stereo_segment)
+    concatenated_audio = np.vstack(concatenated_audio_data)
     concatenated_audio = concatenated_audio / np.max(np.abs(concatenated_audio))
     return concatenated_audio