CNN course: C4W2 PA 1, implementation of ResNet50

Course Q&A Deep Learning Specialization Convolutional Neural Networks
1

On the course about Convolutional Neural Nets I stuble onto an error on 3rd task of
the 1st assignment on 2nd week. Whenever I try to compile all blocks and layers into ResNet50 I get an error after launching the code in the cell where it gets compared with resnet50_summary.

The code of my resnet50 appears to have correct number of lines.
I checked my code in convolutional and identity blocks and it seems to be correct.

Where could the source of the error hide?

AssertionError Traceback (most recent call last)
in
1 from outputs import ResNet50_summary
2 model = ResNet50(input_shape = (64, 64, 3), classes = 6)
----> 3 comparator(summary(model), ResNet50_summary)

/tf/W2A1/test_utils.py in comparator(learner, instructor)
14 def comparator(learner, instructor):
15 if len(learner) != len(instructor):
—> 16 raise AssertionError(f’Models does not have the same number of layers {len(learner)} != {len(instructor)}’)
17 for a, b in zip(learner, instructor):
18 if tuple(a) != tuple(b):

AssertionError: Models does not have the same number of layers 193 != 177

2

Same! I have 176!=177, have no idea where should I focus on.

3

Does this question still need an answer?

4

I got fixed. That’s my stupid… When I was trying to delete one line of comments that I made, I delete one line of code by mistake, which is a layer. That’s the reason why one layer lacks. Thanks for your follow-up.

1 Like
5

Not necessary, sorry for the late reply. For the last week I had a tight schedule but managed to submit the assignment successfully eventually.