Problem in C4W1 programming ex3

The checking section of this exercise suggests that something goes wrong with in this part. But I have thought about this for a long time but failed to figure out the solution. Could you please help me :frowning:

{moderator edit - solution code removed}

Please share your notebook (click my name and message your notebook as attachment)

@TMosh I’m a DLS mentor too. So, there’s no violation.

@Mubsi please fix this. :slight_smile:

That’s odd, because your user name does not have a mentor tag.

Agreed. Mubsi will hopefully fix this.

Hi, Balaji.

As of today, you show up as a Mentor for MLEP. I think Discourse is not swift enough to make your label be context sensitive. E.g. I’m also a mentor for GANs and NLP, but only show up with DLS in my “label” when I post in a different category.

Cheers,
Paul

@paulinpaloalto Thanks for the heads up.

I’d say that the code as shown looks correct to me. What does the grader say? If you miss only 25 points, one possible cause is that you have coerced the type of the output in conv_single_step to be a float32. The unit tests do not catch that, but the grader does.

:grin:Thanks a lot! Understanding that I didn’t make a mistake in this unit is enough for me~

1 Like

Hi @balaji.ambresh,

Please make an issue out of this in the repo and shall take a look at it soon.

Thanks,
Mubsi

There are already several open issues on GitHub about the unit tests for this notebook. The issue with float32 coercion in conv_single_step is this one.

@Mubsi I don’t have access to https://github.com/https-deeplearning-ai/deep-learning-specialization/ or any http-deeplearning-ai repos.