Segmentation fault after killing processes

I am training multiple feedfoward neural networks in parallel via ray. Most of the time the code runs with no problems. Occasionally, though, the code takes a while to run, so I end up killing it. When I do though, I get a segmentation fault when I try to re-run the code. Specifically, I am unable to run any python script where I “import torch”. I am able to execute other python scripts that don’t use pytorch though. I am unsure as to whether this is an issue with my computer, ray, or pytorch.