Comment by getcrunk

I’ve only ran this on my cpu. Using the tiny en model.

Anyone ran tiny en on t4 gpu (aws g4dn iirc)? What’s the speed up