Load VFI Model node to improve efficiency with caching?

[kuprel]

Should there be a separate node to load the VFI model? It seems like film_net_fp32.pt is being loaded every time I run a workflow. For other nodes that require models, e.g. stable video diffusion, stable cascade, the model is only loaded on the first run and then cached.