You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
It seems there is a memory leakage for RecurrentBased type of models RNN, GRU, TCN with Pytorch 2.0 / PL 2.0
It is possible there is more information in the PL github: Lightning-AI/pytorch-lightning#16520
That deprecated the on_validation_epoch_end method.
This issue was partially solved by reducing the memory usage of the models to avoid surpassing RAM.
As the interaction between multiprocess and virtual memory seems to be causing the problem.
The text was updated successfully, but these errors were encountered:
It seems there is a memory leakage for RecurrentBased type of models
RNN, GRU, TCN
with Pytorch 2.0 / PL 2.0It is possible there is more information in the PL github: Lightning-AI/pytorch-lightning#16520
That deprecated the
on_validation_epoch_end
method.This issue was partially solved by reducing the memory usage of the models to avoid surpassing RAM.
As the interaction between multiprocess and virtual memory seems to be causing the problem.
The text was updated successfully, but these errors were encountered: