Grading of ResNet -CNN Week2 assignment

My code for identity block and convolutional block passed the test. Both showed ‘All test passed!’.
But, on submission, the grading showed convolutional block did not pass the test with a note saying:

‘Function convolutional block is incorrect’

Please let me know your suggestion.

Another problem is with ResNet50 part of the assignment. I believe that my coding for the graded function is correct, but when I go to run the code,
model = ResNet50 (input_shape =(64,64,3), classes =6)

I get the UnboundedLocal error showing:

‘local variable ‘identity_block’ referenced before assignment’

Please suggest the solution.


Abdus Shamim Khan

The first issue mostly just means the cases in the public test missed your issue in the code which was caught by grader’s private test. Are you sure you’ve not missed any instructions? You should recheck the instructions and all the passed arguments in the convolutional_block function to see if you’ve used them all properly.

For the second issue, that is indeed weird. I’ll say make sure the order you run cells is correct? I’m not exactly sure about this one though.