Occasionally when I have tried to run a neural network whose parameters are too large on my laptop using TorchStudio or messed around with the dataset analyzers on a dataset that was too large for my computer, TorchStudio GUI would automatically closeout.

However, some of the processes I was trying to run will continue to run in the background. Now, normally this would not be an issue because I could use the top command on my machine to close out any process.

However, when these processes run in the background there are thousands of them and I cannot kill all of them.

Maybe try implementing some additional error messages for when the program reaches resource allocation issues and crashes to prevent this kind of issue.