# Cannot passs week1 Programming Assignment even all tests passed

**URL:** <https://community.deeplearning.ai/t/cannot-passs-week1-programming-assignment-even-all-tests-passed/71522>\
**Category:** Convolutional Neural Networks\
**Tags:** coursera-platform\
**Created:** [December 3, 2021, 7:58pm UTC](https://community.deeplearning.ai/t/cannot-passs-week1-programming-assignment-even-all-tests-passed/71522 "2021-12-03T19:58:24Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![eric\_yang](https://avatars.discourse-cdn.com/v4/letter/e/41988e/32.png) [@eric\_yang](https://community.deeplearning.ai/u/eric_yang)\
**Post date:** [December 3, 2021, 7:58pm UTC](https://community.deeplearning.ai/t/cannot-passs-week1-programming-assignment-even-all-tests-passed/71522/1 "2021-12-03T19:58:24Z")

</div>

for week1 Programming Assignment: Convolutional Model, Step by Step  
I run it in notebook, all tests passed for exercises 1-4, but got 25/100 points?  
anything wrong? or I also need to complete the optional exercises5?

also saw error:  
Tests failed on 3 cell(s)! These tests could be hidden. Please check your submission.

How do I know which test failed and why?

Thanks!

---

<div class="post-metadata">

**Author:** ![paulinpaloalto](https://sea1.discourse-cdn.com/dlai/user_avatar/community.deeplearning.ai/paulinpaloalto/32/187_2.png) [@paulinpaloalto](https://community.deeplearning.ai/u/paulinpaloalto)\
**Post date:** [December 3, 2021, 9:12pm UTC](https://community.deeplearning.ai/t/cannot-passs-week1-programming-assignment-even-all-tests-passed/71522/2 "2021-12-03T21:12:41Z")

</div>

The tests for _conv\_forward_ and _pool\_forward_ are broken in that they don’t catch the bug of omitting the stride in your calculations. So that’s one thing to check. Are you also sure that you didn’t reference any global variables in your “pad” function? It sounds like you failed 3 out of the 4. The other thing to check is to make sure that you didn’t coerce the output of _conv\_single\_step_ to float32. That passes the tests in the notebook, but fails the grader.

Bugs have been filed about all these limitations of the unit tests, but no fixes are available yet.

Note that in the _pool\_forward_ case, there is a _stride = 2_ test case and if your mistake has to do with the stride, some (but not all) of your outputs will be wrong. But no error actually gets thrown in that case, so you just have to look carefully.