Assignment 3: Question Duplicates - Exercise 4: Classify

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

Hello,
When running my Classify function on the Siamese model, I get 13 fewer correct predictions than expected. Out of 10,240 total predictions! How is that possible? Are there rounding errors?

All other exercises in the lab passed ok until this one.

Could a tutor have a look at my code?

Thanks.

1 Like
2

Hi @FranciscoMoreno

Do you have the exact same output?

1 Like
3

Hi @arvyzukai ,
Yes, I do have the exact numbers for accuracy and confusion matrix.
and when I run:

Test your function!

w3_unittest.test_classify(classify, model)

I get:
0 tests passed
8 tests failed

therefore, I can’t submit…

Thanks for your prompt response.

Is possible that the pre-trained model and the unittest are out of “sync”?

1 Like
4

@arvyzukai , any guidance here?

I got 8 test failed. I can’t submit the exercise.

I get the exact same results as other student (you gave me the link) but I don’t know how his issue was solved.

Please, help.

5

Problem solved.
Thanks @jyadav202 for the hint.
The lab is very well explained, it was just me struggling with a step :slight_smile: