C3W2_Compile error during grading

Course Q&A Natural Language Processing NLP with Sequence Models
4

I found two possible problems in my code:

I think the expected “:” could be in the cell #GRADED FUNCTION: predict in the last for loop, which looks something like

for tag_idx in None #HERE IS A : missing!
    pred_label = None
    ...

Regarding why my model did not train, I think the error was due to explicity choosing an axis with tf.reduce_sum() when calculating the masked_acc. But more likely than not, this was just another problem :smiley: Hope this helps!

1 Like