In the first assignment of week 1, i.e., “Python Basics with Numpy”, we have been asked to code the graded function softmax in the seventh part of the exercise. Even though I implemented the 3 steps approach to compute s, I am not getting all the test cases passed. I first applied exponential function element-wise on the input vector/matrix x using np.exp(x). After that, I applied np.sum() to compute the sum row wise. Finally, I used broadcasting while dividing the two variables (vectors/matrices) obtained in the first two steps. Still, it is showing that one test case is getting failed.
I tried printing the outputs of intermediate steps to search where I’m going wrong. But I was unable to figure out. I shall be obliged if someone could help me regarding this. Thank you in advance!
