CNN Week 4 Assignment 1, Excercise 1 Issue with Shape

Course Q&A Deep Learning Specialization Convolutional Neural Networks
1

Good evening. I am working on the first problem of the Week 4 of the CNN course.

I received this outcome:

loss = tf.Tensor(527.2598, shape=(), dtype=float32)

Now, it is not raising an error, but when I turn in the assignment, it is (I believe) counting it incorrect. Is the error because it is not just a singular scalar value? If so, what does the “shape=()” mean and what could be causing this error?

Thank you,
Chris

2

Which function are you working on?
(update - oops, I was looking in Week 1…)
I found it, “triplet_loss()”.

4

Your result is correct. It’s a tensor that is a scalar float - which I interpret to mean that its shape is null.

5

So, that makes sense. However, I am looking at the other excercises, and it seems as if I have them all correct, but I am still getting a 66% on the assignment. The only other thing I can think of is on the third excercise, I recieved the following output:

it’s younes, the distance is 0.5992949
it’s younes, the distance is 0.5992949
it’s younes, the distance is 0.0

Perhaps the last line is incorrect?

Thank you,
Chris

6

Those values are correct.

7

Well, then here are my outputs for Excercise 2:

It’s bertrand, welcome in!
It’s bertrand, welcome in!
It’s not younes, please go away
It’s not younes, please go away
It’s younes, welcome in!

Out[15]:

(0.5992949, True)

If these are correct, then I don’t know why I am only getting 66%

8

All those values are correct also.
Does the grader output tell you which functions didn’t pass?