Grader fails on Classify & Predict

Course Q&A Natural Language Processing NLP with Sequence Models
1

I submitted the assignment and all the unit tests are succeeding. I cannot figure out why the grader is failing on the classify and predict parts (tho the unit tests are succeeding). And it not failing on anything else. The error is a generic “there was a problem, check stderr”. I believe it must be OOMing on the classify call (out of memory).

I check the dimensions of the question matrixes I’m using from the generator and they are (512,128) or (512,64) in the assignment test code.

Could someone take a look at my notebook jbnoaufkielo? I’m hoping someone can tell what’s wrong.

1 Like
2

Hi @adp214

99% of the time when you pass the unit tests (and all your outputs match the expected outputs) but grading of the assignment fails, it is one of those reasons:

Important Note on Submission to the AutoGrader

Before submitting your assignment to the AutoGrader, please make sure you are not doing the following:

  1. You have not added any extra print statement(s) in the assignment.
  2. You have not added any extra code cell(s) in the assignment.
  3. You have not changed any of the function parameters.
  4. You are not using any global variables inside your graded exercises. Unless specifically instructed to do so, please refrain from it and use the local variables instead.
  5. You are not changing the assignment code where it is not required, like creating extra variables.

80% of these are:

  1. You are not using any global variables inside your graded exercises. Unless specifically instructed to do so, please refrain from it and use the local variables instead.

For example, if you have local variable named vocab_size then do not use global variable like len(Vocab), or test_Q1 vs. Q1_test.

Here is a Video what is being meant by global vs. local variables and the playground to experiment.

In any case, make sure you do none of the above 5 points.

Cheers

3

TLDR: I had to put my solution in the original assignment file as the grader ignores the copy file I was working in.

I found the issue, it had nothing to do with the AutoGrader instructions. Might be useful to include this idea in the checklist, especially since this assignment has instructions about copying the file as a precaution when updating to latest version or debugging by restarting the kernel. Could be helpful.

During the debugging I was doing for earlier issues with out-of-memory, I copied the assignment to another file with a different name and continued working in this new file. It appears the “Submit Assignment” button does not submit the current file I am looking at (as I assumed), but rather just submits the original assignment file which remained incomplete for me. I had to copy my code from the file I was working in to the original assignment file and then it worked.