update space

app.py

@@ -153,11 +153,15 @@ def extract_audio(videos_file_path):
 
 @spaces.GPU(duration=120)
 def ola_inference(multimodal, audio_path):
-    visual, text = multimodal["files"][0], multimodal["text"]
-    if visual.endswith("image2.png"):
+    try:
+        visual = multimodal["files"][0]
+    except:
+        visual = None
+    text = multimodal["text"]
+    if visual and visual.endswith("image2.png"):
         modality = "video"
         visual = f"{cur_dir}/case/case1.mp4"
-    if visual.endswith(".mp4"):
+    if visual and visual.endswith(".mp4"):
         modality = "video"
     elif visual:
         modality = "image"