Convolution_model_Step_by_Step_v1 Errors despite it saying "All tests Passed!"

Course Q&A Deep Learning Specialization Convolutional Neural Networks
1

Greetings,

This is regarding the first assignment in week 1 of course 4… My “conv_forward(A_prev, W, b, hparameters)” seems to be correctly coded but despite that, the results are different and the notebook gives me a “All tests passed!” comment.

This is weird as well when going into the Pooling stage because when stride=1, the result is correct. But when stride=2, then something is not right.

Please help me. Thanks in advance.

2

Hi @Omar_Said_Brito_Sala,

When you are looping over the batch of training examples, particularly “loop over vertical axis of the output volume”, you are not considering stride into account. This is why when the default value of stride is 1. you are able to pass, but when the stride value changes in the unit test, you fail.

Hope this helps,
Mubsi

3

Thanks for the tip. I considered the stride in my code and it worked. Thanks again!

1 Like