C3W3 classify all test passed but when graded, the score is 0

[skinx.learning]

In the assignment is all test passed the output is matched but when grading my score is zero.

[Deepti_Prasad]

Click on the show grader output for classify section to know why it failed

[skinx.learning]

It’s show

There was an error grading your submission. Details:\n{{function_node _wrapped__Mul_device/job:localhost/replica:0/task:0/device:CPU:0}} Incompatible shapes: [3500,64] vs. [3500,0] [Op:Mul] name:

I print all the shape that in function classify but has no shape of [3500,64] or [3500,0].

[Deepti_Prasad]

Hi @skinx.learning

remember the difference between the incompatible shape is related your second dimension that is 64.

Can I know how are you recalling your pred in the classify grader cell?

I suspect issue with the below

pred =
_, n_feat =
v1 =
v2 =

C3 W3 Ex4 _, n_feat values

Hi @Lujain_Andijani There’s a hint couple of cells above, in the TripletLoss() function implementation: def TripletLoss(labels, out, margin=0.25): _, embedding_size = out.shape # get embedding size ... Note, that in your case the variable is not named out, but pred. So, to get the dimensionality (n_feat) of the model’s prediction, you just need to use the “shape” property of this variable. Cheers

Extra hint

Regards
DP

[skinx.learning]

pred = {codes removed by moderator}
_, n_feat = (codes removed by moderator}
v1 = pred[:, :128] incorrect code
v2 = pred[:, 128:] incorrect code

Is anything wrong ?

[Deepti_Prasad]

@skinx.learning

v1 = pred[:, :128]
v2 = pred[:, 128:]

You are suppose to define the v1 and v2 according to _, n_feature and not assign values directly.

Refer

@Deepti_Prasad

this image and do the changes

[skinx.learning]

Problem Solved!! Thanks for your help.