Of course. Problem is, a lot of the kernel halt requests don't seem to make it 90% of the time and restarting the kernel gets annoying when using it on larger datasets. Though, it has forced me into writing easily re-runnable code which is kinda good.