Could you please confirm whether the inference‑mode parameters provided in the repository are correct? Why do I only get noisy videos after running inference?

[meteorite2023]