Why do you need to divide by 255 when using stack function?

Part 1 (2020)
1

I’m trying to understand the concept behind dividing by 255 when you call the stack function when placing all images into a rank-3 tensor. This is related to the 04_mnist_basics code.

stacked_sevens = torch.stack(seven_tensors).float()/255
stacked_threes = torch.stack(three_tensors).float()/255

There was no explanation as to why the number to divide by is 255 and I can’t figure out where the math for this comes into play.

2

found the answer to my own question.

As the pixel values range from 0 to 256, apart from 0 the range is 255. So dividing all the values by 255 will convert it to range from 0 to 1.

1 Like