Yesterday, I started a run which I terminated it after a few minutes of running. But until now, it keeps running and does not allow me to terminate again, stating the user had already terminated. However, it is blocking me to do any other submission.
I couldn’t find the root cause, so I forced termination.
Also, I will try to come up with a more permanent solution for:
AttributeError: 'tqdm_notebook' object has no attribute 'disp'
as you are not the only one who has had this problem.
An easier way is to do this:
from tqdm.auto import tqdm`
First thing I did on the next run I would try, was for sure, removing tqdm haha @enzo and my guess is that when using the .notebook one, you must have ipykernel, which does not seem to be on the env.
When you submit a notebook, it is converted to a python script file to make sure it will always be parsed the same and you can verify the result of this conversion.
If you want to know all the thing that are being processed, please have a look at the documentation.
@enzo, I might need your help again. I’ve tried the run with the change #38697. However, it is giving an error on the infer step, requesting to connect with support.
Also, the first one that was not terminating, consumed almost all of my available quota. Is there any way to get those back sooner?
Failed runs do not count toward the quota. This is to avoid situations where you time out and have to wait until next week.
I just gave you 10 extras hours as this is the first week.
@enzo Thnak you! Could you also help me understand why the run #38697 failed on the infer step? I am not being able to understand from the log. What is a bit weird is that the code ran on colab.
That’s what I was talking about.
Since you only had 1h30mins -and we always give an extra 30mins- you timed out after 2h30mins.
Just re-run it as you got more quota now.
Oh, now I got it! sorry.