fix

app.py

@@ -351,7 +351,7 @@ def run_diffusion(audio_latent: torch.Tensor, meta: dict, duration: float) -> to
 
 # ==================== Full Inference Pipeline ====================
 
-@spaces.GPU
+@spaces.GPU(duration=120)
 def generate_audio_core(video_file, caption):
     
     global DEVICE
@@ -535,23 +535,17 @@ def build_ui() -> gr.Blocks:
                 gr.Markdown("### 📋 Run Log")
                 log_output = gr.Textbox(
                     label="",
-                    lines=18,
-                    max_lines=30,
+                    lines=10,
+                    max_lines=15,
                     interactive=False,
                     elem_classes=["mono"],
                 )
-
-        # ======================================================
-        # Row 2 — Output video (original video + generated audio)
-        # ======================================================
-        gr.Markdown("---")
-        gr.Markdown("### 📤 Output  —  Original video with generated audio")
-
-        video_output = gr.Video(
-            label="Video + Generated Audio",
-            interactive=False,
-            height=400,
-        )
+                gr.Markdown("### 📤 Output")
+                video_output = gr.Video(
+                    label="Video + Generated Audio",
+                    interactive=False,
+                    height=300,
+                )
 
 
          # ======================================================