Added Cumprod to raw_ops

[KareemMAX]

Close #4944

[KareemMAX]

• Cumprod #4944

[KareemMAX]

Can you please review my PR when you have a chance? @abdrahmandiab

[KareemMAX]

@abdrahmandiab I noticed you made some changes, can help me understand it? And I was wondering how the review is going?

[abdrahmandiab]

@abdrahmandiab I noticed you made some changes, can help me understand it? And I was wondering how the review is going?

Hi @KareemMAX , I believe I made 2 changes
first, to the tests, I removed the tolerance values rtol & atol since they are limiting failure detections and are unnecessary as tests pass without them. I also removed the with_out argument generation since for a Tensorflow backend, it is always set as False.

and second, I added aliases for cumprod and cumsum in tensorflow/math.py. This is to match the Aliases used by Tensorflow since it has a tf.math.cumprod and a tf.raw_ops.Cumprod, and same goes for cumsum (which is not really related to your PR 😅 )

I will be adding some review comments soon to address a few tiny things that have to be updated so stay tuned for that!

[abdrahmandiab]

Hi @KareemMAX , great work so far! We are very close to merging this PR. 🚀

Just a few changes outlined in this review that hopefully should not take too much of your time!

It would also be nice if you handle the merge conflicts currently present with the branch, and let me know if you need any help with that!

Thanks! 😇

[KareemMAX]

Hey @abdrahmandiab, thank you so much for the insights.
I have addressed your comment, while doing so I have found some bugs in cumprod backend which I fixed and tests now pass. Awaiting for your review.

Thanks a lot! ❤

[abdrahmandiab]

Hi @KareemMAX,
Could you give me an example that was failing due to the bug you fixed? :)

Otherwise, everything looks great!
Thanks so much for your work, will be merging soon! 🚀

[KareemMAX]

Hey @abdrahmandiab,
When exclusive=True tests fail, this was due to backends multiplying tensors with zeros_like(x), I changed them to ones_like(x) while keeping tensorflow backend as is. You can also refer to tensorflow api reference to recheck my changes.

Thanks ❤