In config.py:
parser.add_argument('--padd_size',type=int,help='net pad size',default=0)#math.floor(opt.ker_size/2)

I'm wondering why you commented out the code for same padding to use valid padding instead? Is this important? (I checked both paper and ICCV talk, and it wasn't mentioned)

Can you please add some performance numbers to the main project docs indicating inference latency running some common hardware options e.g. AWS p2, GCP gpu instance, CPU inference, Raspbery pi, etc.

Description

In the changelog it is stated that there is a linux port..... However that is absolutely nowhere to be found, not on the wiki, and nowhere on the github page, what gives?

Also, i know telegram would be more of an appropriate place for this, but i don't have a carrier number as i don't need one and honestly cannot be bothered to bug someone else to use their carrier number sim

Is there any documentation available for the code? At least, information about what each function does? It's really difficult to read through the entire 1000+ lines of code without being able to grasp which function is called where.