C2_W4 assignement "Run the distributed training in a loop"

Hi everyone,

I have been trying to do the C2_W4 assignment on distributed training loops.
I tried to run the training loop on Coursera Notebook.
When I run the cell, nothing appears, I just see 0it [00:00, ?it/s] in the output.
I let it run for a while and when I came back the kernel had died.
I am not sure what is happening, the previous parts of the notebook seem to be running well. Anybody has a clue on that?

Thanks a lot

Hey!
Have you tried restarting the kernel? Also, make sure that at least 20 minutes have passed after running the cell, as stated in the comments.
Additionally, double-check that the previous codes are correct, as any errors in preceding cells could potentially affect the execution of the current cell.
Let us know if you need further assistance!

Hello @simon3

Can you once confirm, if you have not edited any of the non-grader cells? And/or added any extra statements other than where code writing was not required and/or deleted any code cell?

This kind of issue happens when the model cannot be trained due to issue in codes too, cross check again if you have not added any extra code than asked.

Regards
DP

Hello,

Thanks for your advices!
Now it works. I think the problem was that I was using loss_object instead of compute_loss to compute the loss of the test sample.
But I had no warning before training the model that something was wrong.

Cheers,
Simon

1 Like