Assigment 1: log_perplexity: There was an error grading your submission. Details: operands could not be broadcast together with shapes (2,7) (2,)

[Deepti_Prasad]

Hi @Erika_Petersen

Kindly refer the below comment by @arvyzukai to check if you have passed the correct axis value of -1.

ValueError: operands could not be broadcast together with shapes (2,7) (2,)

Hi @Cawnpore_Charlie This unit test is a bit contrived (most probably accidentally, but also could be intentionally). Let me explain the last test case (2,1,7,3) that you’re having problems with. (Usually, this is not a normal input shape in NLP but it is used in the test case). So, in the unit test case the preds are: <tf.Tensor: shape=(2, 1, 7, 3), dtype=float32, numpy= array([[[[0.1 , 0.5 , 0.4 ], [0.05, 0.9 , 0.05], [0.2 , 0.3 , 0.5 ], [0.1 , 0.2 , 0.7 ], …

Let me know if it still confuses you, you can send a screenshot of the log_perplexity codes via personal DM.

Regards
DP