MPS device compatibility

[adills]

With Pytorch and TF both supporting Apple's M1/M2 silicon chips, you should be able to include a device checker and include "mps" capability for GPU processing. If only I had the time to create a branch and show you.

[Spycsh]

Thanks for info:) I do not have access to a Mac currently and the optimization here in my repo should only be used for intel xeon cpu. But good to know about the mps device compatibility. Is it fast enough?

[adills]

Thank you for your quick reply. I can run inference.py using your int8 instructions and --cpu on my MacBook M1 Max Pro but at a rate of ~18s/it (not even 1 it/s). If I get some time, I might create a fork and see if I can modify your engine to accept device=mps.