DLS/Course 4/Week 2/CUDA runtime implicit initialization on GPU:0 failed. Status: out of memory

labID : apmdgplf GPU error

Restarting the server and "get latest version" may only be useful because they have a chance of assigning you a different GPU. See the following replies for more information.

Thanks for reporting. I confirm that the problem is related to the infrastructure and not to your notebooks. We use shared GPUs for the notebooks, and we cannot limit the amount of memory that each process can use. The only workaround is that each learner stops the notebook once finished, for releasing the resources and collectively reduce the odds of having the issue. Nothing else we can do for now. Just keep learners informed.

So, there’s no way to proceed with the course itself, sinceit’s impossible to submit tasks?

I am facing with the same issue here, so how we are expected to proceed with this assignment? I did rebooting and it did not give me a solution. I cannot move forward

Updated 7/28/2022:
Try restarting the server (Help → Reboot Server). The issue appears to be a hardware problem in Coursera’s GPU array. Restarting the server may assign you a different GPU.

I couldn’t find Reboot Server button
image
So i fixed this by referencing H​ow to Refresh your Workspace in course 5.
I appended "?forceRefresh=true " to assignment URL and it started to work.

It’s in the other “Help” menu.

Just to let you know that this issue still exists as of 8/9/22. It took several reboots to get it to work. For all those times, I thought I was doing something wrong (even though it worked on the lab files I downloaded to work in my own environment).