Having multiple kernel crashes

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

Hi there,
On the accuracy calculation, I have multiple kernel crashes even though I’m running only 1 notebook. Can you please provide some tips to get it to run?

Thanks,
Rama

2

Hey @Rama_Mahajanam,
If I am not wrong, you are talking about the classify function, right? Nonetheless, it’s strange that you are facing kernal crash specifically on this exercise. Most probably, you might have written some code that consumes a lot of memory and compute, and ultimately crashes your kernel.

Can you please DM me your code for the same, so that I can ensure that it’s not the code, but a Coursera issue, in which case, you can post an issue in Courser’s Help Centre.

Cheers,
Elemento

3

yes, it’s the classify function. I can DM you the code

4

Sure @Rama_Mahajanam,
For DM, you can click on my name and select “Message”.

Cheers,
Elemento

5

thanks, just did that :slight_smile:

6

Hey @Rama_Mahajanam,
In the very first line of code itself of the classify function, you have called data_generator with incorrect arguments. Since you haven’t passed arguments in the form of parameter_name: parameter_value, it took vocab['<PAD>'] as the batch_size, and vice-versa. Either ensure that your arguments are passed in the correct order, or use the aforementioned convention. As a result, it is taking a lot of memory and your kernel is crashing.

Additionally, your code for computing the accuracy is also incorrect. Please read the instructions carefully and correct it accordingly. I hope this helps.

Cheers,
Elemento