configuration for Macs(apple silicon)

[bawsi99]

I found having the following configuration works best to unleash the full potential of the gpu available in any apple silicon .
just add the following piece of code to train.py :

import torch._dynamo
torch._dynamo.config.suppress_errors = True
device='mps' (set device to mps), I have added the image of cpu and gpu utilisation of my machine from running this code .