C4W1_UNQ_C4: UNIT_TEST Failed - Why?

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

The function NMTAttn() generates a model that matches the expected model output, when visually inspected. Here are the two files
expected.py (721 Bytes)
generated.py (714 Bytes)
However, the w1_unittest.test_NMTAttn(NMTAttn) code reports:
6 Tests passed
2 Tests failed

How can I troubleshoot it, without knowing what part of the code failed the test?

2

Never mind. I found the mismatch on the last line :slight_smile: I was supposed to use LogSoftmax instead of Softmax function

3

Hi @mwcruz,

Glad you were able to fix it on your own.

Best,
Mubsi