I want to ask for the the first week’s programming assignment —— Convolutional Model, Step by Step of Convolutional Neural Networks.
The outcome of my code is diffierent from the right answer. How can I solve this? Thanks.
Hi,
First, you shouldn’t share your code. Kindly delete it.
Second, you just need to follow the give instructions:
# Element-wise product between a_slice_prev and W. Do not add the bias yet.
# s = None
# Sum over all entries of the volume s.
# Z = None
# Add bias b to Z. Cast b to a float() so that Z results in a scalar value.
# Z = None
You have to remove the None
and write the correct code there. Don’t need to write anything at other places. So, you just have to write the code for s
, Z
, and again, Z
.
Hint:
- Element-wise product between
x
andy
is:x*y
- Check numpy sum for summing over all entries of the volume s
- Adding
x
andy
is:x + y
. Buty
should be float. Check on Google aboutfloat()
.
Best,
Saif.
Another thing to notice about the initial implementation is how many times b was being added because of being included within loop on i. Compare that to the guidance provided, where b is added once at the end.