How to count viterbi_forward?

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

I have output of the test like

Wrong values for best_probs.
Expected: [[-22.60982633 -24.78215633 -34.08246498 -34.34107105 -49.56012613]
[-23.07660654 -24.51583896 -35.04774303 -35.28281026 -50.52540418]
[-23.57298822 -29.98305064 -31.98004656 -38.99187549 -47.45770771]
[-19.76726066 -25.7122143 -31.54577612 -37.38331695 -47.02343727]
[-24.74325104 -28.78696025 -31.458494 -36.00456711 -46.93615515]].
Got: [[-22.60982633 -24.78215633 -18.63206488 -18.63206488 -18.63206488]
[-23.07660654 0. 0. 0. 0. ]
[-23.57298822 0. 0. 0. 0. ]
[-19.76726066 0. 0. 0. 0. ]
[-24.74325104 0. 0. 0. 0. ]].
Wrong values for best_probs.
Expected: [[-202.75618827 -208.38838519 -210.46938402 -210.15943098 -223.79223672]
[-202.58297597 -217.72266765 -207.23725672 -215.529735 -224.13957203]
[-202.00878092 -214.23093833 -217.41021623 -220.73768708 -222.03338753]
[-200.44016117 -209.46937757 -209.06951664 -216.22297765 -221.09669653]
[-208.74189499 -214.62088817 -209.79346523 -213.52623459 -228.70417526]].
Got: [-22.60982633 -23.07660654 -23.57298822 -19.76726066 -24.74325104
-35.20241402 -35.00096024 -34.99203854 -21.35069072 -19.85767814
-21.92098414 -4.01623741 -19.16380593 -21.1062242 -20.47163973
-21.10157273 -21.49584851 -20.4811853 -18.25856307 -23.39717471
-21.92146798 -9.41377777 -21.03053445 -21.08029591 -20.10863677
-33.48185979 -19.47301382 -20.77150242 -20.11727696 -20.56031676
-20.57193964 -32.30366295 -18.07551522 -22.58887909 -19.1585905
-16.02994331 -24.30968545 -20.92932218 -21.96797222 -24.29571895
-23.45968569 -22.43665883 -20.46568904 -22.75551606 -19.6637215
-18.36288463].
Wrong values for best_paths.
Expected: [[ 0 11 20 25 20]
[ 0 11 20 25 20]
[ 0 11 20 25 20]
[ 0 11 20 25 20]
[ 0 11 20 25 20]].
Got: [[ 0 11 33 33 33]
[ 0 0 0 0 0]
[ 0 0 0 0 0]
[ 0 0 0 0 0]
[ 0 0 0 0 0]].
Wrong values for best_paths.
Expected: [[20 19 35 11 21]
[20 19 35 11 21]
[20 19 35 11 21]
[20 19 35 11 21]
[35 19 35 11 34]].
Got: [[0 0 0 0 0]
[0 0 0 0 0]
[0 0 0 0 0]
[0 0 0 0 0]
[0 0 0 0 0]].
2 Tests passed
4 Tests failed

What can be a problem?

2

Hi @someone555777

Actually, first you should check if you have not hard-coded any values:
def viterbi_forward(A, B, test_corpus, best_probs, best_paths, vocab, verbose=True):

You should use these variables in your implementation and not other.

Cheers

3

I don’t have ideas. Can you check assigment in your private messages?

4

You have messaged me a link to the Coursera webpage :slight_smile:

5

Just click on it :slight_smile: looks like all links transform to images

6

You should know that clicking the link you sent me would redirect me to my assignment (with the help of cookies)
Anyway, I don’t want to get into explaining how internet works :slight_smile: but the link you’ve sent me is not what you think - for me it does not lead to your Assignment (but for you it does).

7

Also, have you checked if you use A, B, test_corpus, best_probs, best_paths, vocab variables exclusively (case sensitive, these variables and not any others)?

9

For future readers, the OP’s problem was in defining both for loops’ ranges:
# For each unique POS tag that the current word can be
# For each POS tag that the previous word can be:

In this Assignment the j and k values (ranges) should be bounded by num_tags and not some other strange number.