-
In chap 13 ( convolutions ) there is a line that say: “it will return a1+a2+a3-a7-a8-a9. if we …” Should it be “-a1-a2-a3+a7+a8+a9” ? Plus, same sentance, " if a1 is greater than a7, a2 … get a bigger number as results". should it be “smaller number” ?(edited)
-
In the book ( also in chap 13 ), 255 is shown as black, and the document talks about edge finding based on the color, while the conversion for white is 255 and black as 0. Any reason for this “flip”?
This one is not a question, but rather an error in the book:
Original: " Kernels passed to F.conv2d need to be rank-4 tensors: [channels_in, features_out, rows, columns]"
Should be: " Kernels passed to F.conv2d need to be rank-4 tensors: [ out_channels,in_channels, rows, columns]"