没有看到inference代码

[Harold0530]

这是一篇非常好的工作，我有个疑问是目前好像就是对整个数据集检索+生成，好像没看到单个的prompt+image的输入，然后去检索生成吧
python main.py test -c configs/cogvideox/MotionRAG_open.yml

[flateon]

Thanks for your interest in our work!
Currently, we only provide dataset-based retrieval and generation in the code release. However, we are planning to release a Gradio demo in the near future, which will allow you to input a single prompt and image for retrieval-augmented video generation. Please stay tuned!

感谢您对我们工作的关注！
目前代码版本中我们仅提供了基于数据集的检索和生成功能。不过，我们计划在不久的将来发布一个 Gradio 演示界面，届时您可以输入单个提示词和图像来进行检索增强的视频生成。敬请期待！