forgot to load model

app.py

@@ -66,6 +66,7 @@ def inference(
 ):
     # default processor
     processor = AutoProcessor.from_pretrained("Qwen/Qwen2.5-VL-7B-Instruct")
+    model = load_model(use_flash_attention=True)
     fps = get_fps_ffmpeg(video_path)
     logger.info(f"{os.path.basename(video_path)} FPS: {fps}")
     messages = [