Update app.py

app.py

@@ -83,7 +83,7 @@ def predict_answer(image, video, question, max_tokens=100):
         frames = video_to_frames(video)
         answers = []
         for frame in frames:
-            image = Image.open(extract_frames(frame)).convert("RGB")
+            image = extract_frames(frame)
             image_tensor = model.image_preprocess(image)
             
             # Generate the answer