>>105749147
>You should probably be using Lightx2v instead.
I am. The rentry says FusionX should be a "drop-in replacement" for the default Wan Model when using the Lightx2v workflow, and mentions that Lightx2v should be dropped to 0.6 strength.
I have it set up like so.
FusionX gguf> Lightx2v (@0.6str)> General loras> PatchModelOrder> TorchCompileModelWanVideo> WanVideoNAG > Apply RifleXRoPE WanVideo

Here's a screenshot of the main bit of the workflow. Maybe someone can glean something from it. Not pictured are the video combine nodes.