Week 3 Exercise 8 Wrong Value of W1

Course Q&A Deep Learning Specialization Neural Networks and Deep Learning
1

Cost Value is not changing. I have checked for hard code but not found any.
None of the values of W1, W2, b1 etc matching
please help

Cost after iteration 0: 0.689099
Cost after iteration 1000: 0.689099
Cost after iteration 2000: 0.689099
Cost after iteration 3000: 0.689099
Cost after iteration 4000: 0.689099
Cost after iteration 5000: 0.689099
Cost after iteration 6000: 0.689099
Cost after iteration 7000: 0.689099
Cost after iteration 8000: 0.689099
Cost after iteration 9000: 0.689099
W1 = [[-0.00643025  0.01936718]
 [-0.02410458  0.03978052]
 [-0.01653973 -0.02096177]
 [ 0.01046864 -0.05990141]]
b1 = [[-1.02420756e-06]
 [ 1.27373948e-05]
 [ 8.32996807e-07]
 [-3.20136836e-06]]
W2 = [[-0.01041081 -0.04463285  0.01758031  0.04747113]]
b2 = [[0.00010457]]
---------------------------------------------------------------------------
AssertionError                            Traceback (most recent call last)
<ipython-input-55-6d3d4c22ad68> in <module>
      7 print("b2 = " + str(parameters["b2"]))
      8 
----> 9 nn_model_test(nn_model)

~/work/release/W3A1/public_tests.py in nn_model_test(target)
    273     assert output["b2"].shape == expected_output["b2"].shape, f"Wrong shape for b2."
    274 
--> 275     assert np.allclose(output["W1"], expected_output["W1"]), "Wrong values for W1"
    276     assert np.allclose(output["b1"], expected_output["b1"]), "Wrong values for b1"
    277     assert np.allclose(output["W2"], expected_output["W2"]), "Wrong values for W2"

AssertionError: Wrong values for W1
2

Inside the loop in nn_model, are you calling update_parameters after computing gradients?

3

Yes, I am calling update_parameters after computing gradients.

4

Are you calling optimize inside the model function?

5

There is no optimize function in the Exercise

6

Sorry. Wrong window. Please message me your code. Click on my name and Message to attach your code.

7

The problem was fixed. There was a minor bug wrt assignment.

8

Facing the same issue at my end. Tried many different things, but all steps above seem to work. It simply fails at the nn_model with the attached error.

image
image1758×1023 124 KB

9

Welcome to the community.

If none of the above solves your case, it is better to create a new thread.
In you case, bias is always 0. So, most likely update_parameters() is not called or failed.
This should be the starting point. Please double check.

10

Created a new thread and was able to resolve it after a few suggestions from Paul. Thanks @anon57530071 for responding!