i am currently getting this error
screenshot removed as it contained grade cell codes and posting codes correct or incorrect is against community guidelines
hi @arungnawali
Arguments of grade cell propagate mentions
w -- weights, a numpy array of size (num_px * num_px * 3, 1)
X -- data of size (num_px * num_px * 3, number of examples)
so your error output is mentioning the activation codes do not match with dimensions.
Thinks to check, in your previous grade cell where you initialise the parameters with shape, you need to check if the weight was correctly initialised with vector of shape (dim, 1) as well as for bias as using 0 would create an integer and 0. would create a floating number.
Regards
DP
i have checked every thing is correct and upto exercise 5 every test are passed but unable to pass exercise 6 due to broadcasting issue while computing A
Can DM screenshots of previous grade cells codes. Click on my name and then message.
I think I’ve seen that (2, 3) vs (2, 2) error caused by mistakes in the “update parameters” logic. E.g. using dw to update b.
Note that the bug is in optimize, not in propagate. Note that nothing in propagate changes the value of X or w, so they were already wrong at the point that propagate was called.
1 Like
hi @arungnawali
In your grade cell initialise with zeros, you seem to have edited the grade cell beyond the assigned markers ###YOUR CODE START HERE and YOUR CODE ENDS HERE###
your codes are correctly written but you have edited or deleted
assertion statement already given when you open the assignment notebook
assert(w.shape == (dim, 1))
assert(isinstance(b, float) or isinstance(b, int))
So my advise would be first get a fresh copy of assignment and re-write codes only between the assignment markers and without deleting any other part of instructions or already provided codes for successful submission.
regards
DP
1 Like