Update app.py

app.py

@@ -560,17 +560,19 @@ def get_duration(image_path, audio_path, text, orientation_state, num_steps, ses
 
         return 0
 
+    if session_id is None:
+        session_id = uuid.uuid4().hex
 
-    # audio_chunks = inferpipe.get_times(
-    #             prompt=text,
-    #             image_path=image_path,
-    #             audio_path=audio_path,
-    #             orientation_state= orientation_state,
-    #             seq_len=args.seq_len,
-    #             num_steps=num_steps
-    #         )
+    limited_audio_path = preprocess_audio_first_5s_librosa(audio_path, True, session_id)
 
-    audio_chunks = 1
+    audio_chunks = inferpipe.get_times(
+                prompt=text,
+                image_path=image_path,
+                audio_path=limited_audio_path,
+                orientation_state= orientation_state,
+                seq_len=args.seq_len,
+                num_steps=num_steps
+            )
     
     warmup_s = 30
     duration_s = (20 * num_steps) + warmup_s