Inconsistent Scoring

This happens in every submission. Some step doesnt get the score and I always end up with 80 to 85!! Look at the attached screen shot - I did the data channels dict set up. Logically if this is not done, one can’t proceed, right? Dont really know what is expected and how this scoring stuff happens

Which lab was this for? Also, if you can provide a screenshot of where it thinks the error is would be helpful.
Thanks,
Graham

Hi. This was for class 3 week 3. I attach the screenshot in the original note. I’ve actually completed the course, but just wanted to let you guys know that there’s something messed up with the scoring system.

I experienced the same problem with C3_W1_Assignment. The fist exercise is scored at 0/15 while others are scoped at 100%. If the first assignment of ‘train’ and ‘validation’ datasets would be wrong the next steps should fail as well. I believe it’s about scoring the code below (answers atr left out)

from sagemaker.inputs import TrainingInput

data_channels = {
### BEGIN SOLUTION - DO NOT delete this comment for grading purposes
‘train’: ANSWER_GOES_HERE, # Replace None
‘validation’: ANSWER_GOES_HERE # Replace None
### END SOLUTION - DO NOT delete this comment for grading purposes
}

I had a similar observation, ‘train’:TrainingInput(s3_data=processed_train_data_s3_uri), tried saving and submitting couple of times, it was able to grade the test 1 correctly.

I had the same problem, very frustrating to not get the correct grade.