How fast is the inference?
wtesler opened this issue · 1 comments
wtesler commented
Before I get too far into this project, I'm interested in knowing how fast this works. Is it even remotely close to realtime? Or is it multiple seconds per frame?
Tetsujinfr commented
Although they show an animated gif with a smartphone "demo" of possible effects on the readme intro, lower on the readme they mention processing a 3s video takes 37mins on their 2080RTX system: haha good luck for realtime folks! the ML bs at its best!