Why do I have an 88/100 in coursera if in my jupyter notebook I passed all the tests?
The tests in the notebook to not prove that your code is completely correct. There are lots of ways to write code that passes one test, but not another.
But it is not very helpful for the grader not to tell us where to look or even which function(s) failed. We probably need to look at your whole notebook. I’ll send you a DM about how to do that.
To close the loop on the public thread, the problem was that the code in nn_model was accidentally overriding the value of n_h to always be 4, instead of the value that was passed in as a parameter. The unfortunate thing is that the tests in the notebook do not catch that error, but only the grader does.
I’ll file an enhancement request to suggest modifying the tests in the notebook to catch this error.
2 Likes