chat,
someone said if you use fusionX model with self-forcing lora, the motion is smoother.

Which one of these should I add to my regular WAN+light2x workflow to create similar effect though? and which one to avoid
I don't want to use fusionX