RuntimeError: One or more background workers are no longer alive. Exiting. Please check the print statements above for the actual error message

[RunChuan123]

When I processed the data for the first time and was ready to run, it reported this error. When I changed to a100 training on colab, the memory would always increase inexplicably, causing colab to crash. It even crashed and disconnected before the error message was printed. I have tried many times. I hope someone can help me.

[maxencewynen]

What is the command that prompted the error? Was it during planning, preprocessing or training? Do you have any logs to show associated to this?
It does look like the crash is due to your CPU RAM being overloaded. I would suggest making sure you are running critical code on your GPU and not your CPU.