Week 1 Dinosaur Assignment Ex.1

Course Q&A Deep Learning Specialization Sequence Models
1

Hi,
It seems like the default code has some issues with one of the for loops. Before, I was getting this error

TypeError: builtin_function_or_method is not iterable Solution

for this for loop:

for grad in gradients2.keys:
valuei = gradients[grad]
valuef = gradients2[grad]
mink = np.min(valuef)
maxk = np.max(valuef)
assert mink >= -abs(mValue), f"Problem with {grad}. Set a_min to -mValue in the np.clip call"
assert maxk <= abs(mValue), f"Problem with {grad}.Set a_max to mValue in the np.clip call"
index_not_clipped = np.logical_and(valuei <= mValue, valuei >= -mValue)
assert np.all(valuei[index_not_clipped] == valuef[index_not_clipped]), f" Problem with {grad}. Some values that should not have changed, changed during the clipping process."

I had to change it to keys() to make it work.

If this fix is not supposed to work - then perhaps I wonder why .keys should be correct? Thanks

2

Do not modify the code in the test functions. Doing that just masks errors in your function code.

3

Or, you need to restart the kernel and run the notebook cells again.

4

Hi, maybe you can enlighten me as to why i was getting an error. I tried refreshing the kernel but it didn’t work.

with .keys, the error reads the following:

Gradients for mValue=10
gradients["dWaa"][1][2] = 10.0
gradients["dWax"][3][1] = -10.0
gradients["dWya"][1][2] = 0.2971381536101662
gradients["db"][4] = [10.]
gradients["dby"][1] = [8.45833407]
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-6-73d1a4f3ccd5> in <module>
     29     print("\033[92mAll tests passed!\x1b[0m")
     30 
---> 31 clip_test(clip, 10)
     32 clip_test(clip, 5)
     33 

<ipython-input-6-73d1a4f3ccd5> in clip_test(target, mValue)
     17     print("gradients[\"dby\"][1] =", gradients2["dby"][1])
     18 
---> 19     for grad in gradients2.keys:
     20         valuei = gradients[grad]
     21         valuef = gradients2[grad]

TypeError: 'builtin_function_or_method' object is not iterable

My clip function works fine in the later sections of the assignment. I know that I can’t post the exact code here, but basically what I did is to iterate through a list of the exploding gradients as suggested, and then used numpy.clip. I used the out= parameter.

As far as I understand, the function generates a dictionary object. The .keys() method returns the keys in a dictionary object. I dont think .keys works for dictionaries.

5

What exactly do you mean by “with .keys”?

6

You can see that in the error message. The original test function has the following for loop:

for grad in gradients2.keys:

“.keys” refers to this call of an attribute called “keys” - but it does not seem like that is how you would call the keys in a dictionary…

7

This is how it looks in my copy of the notebook:

image
image880×607 39.8 KB

I think maybe your notebook was accidentally edited to remove the ().

8

OK Thanks. Glad to know that is how it’s supposed to be.