dreamquark-ai / tabnet

PyTorch implementation of TabNet paper : https://arxiv.org/pdf/1908.07442.pdf
https://dreamquark-ai.github.io/tabnet/
MIT License
2.61k stars 485 forks source link

AssertionError: can only test a child process #433

Closed sonnguyen129 closed 2 years ago

sonnguyen129 commented 2 years ago

Describe the bug

Sometimes when I trained that model, after some epochs I cause that error

What is the current behavior?

If the current behavior is a bug, please provide the steps to reproduce.

Expected behavior

Screenshots

image

Other relevant information: poetry version:
python version: Operating System: Additional tools:

Additional context

Optimox commented 2 years ago

Maybe you can try to set num_workers=0

Optimox commented 2 years ago

@sonnguyen129 do you have more information on your issue ? Because I don't think I can help with so little information.

sonnguyen129 commented 2 years ago

Hi @Optimox, I think I solved it, just set num_worker=0 as you said. Can you explain that warning? Thank you so much

Optimox commented 2 years ago

num_workers is the number of process trying to read lines of the matrices in the dataloader. If it's set too high for your computer configuration I guess there can be issues with parallel computing.

Glad that your problem is solved.