There is an issue with the shape of the input in a test case.
I have attached the screenshot.
{moderator edit: code removed}
Your screen image does not appear to show any problems with the shape.
Do you have some test results you can post?
I recommend you read all of the “Additional Hints for unrolling” from the instructions just above that code cell. I think you may be missing a few steps.
Also, please don’t post your code on the forum. That isn’t allowed by the community guidelines.
I have removed the code from your post.
the highlighted part is the skeleton code provided in the assignment.
I think for the first test case, when I call the shape function to get the shape of a_G I get a tuple of length 4 which is expected, and for the second test_case I get a tuple of length 2 instead of 4
as you can see in the first image
Did you review the instructions that I mentioned? Your code seems to be missing the transpositions.
We’ve used this test case for about two years. It is unlikely you discovered a previously unknown problem.
Why you are using 1 while reshaping a_C_unrolled and a_G_unrolled?
See the hint:
To unroll the tensor, you want the shape to change from (m,nH,nW,nC) to (m,nH×nW,nC) .
tf.reshape(tensor, shape) takes a list of integers that represent the desired output shape.
For the shape parameter, a -1 tells the function to choose the correct dimension size so that the output tensor still contains all the values of the original tensor.
So tf.reshape(a_C, shape=[m, n_H * n_W, n_C]) gives the same result as tf.reshape(a_C, shape=[m, -1, n_C])
Best,
Saif.