Page 1 of 1

Tensorflow error happens randomly while training

Posted: Tue Oct 17, 2023 9:57 pm
by semaj4712

I am getting the following error

2023-10-15 12:28:59.400559: F tensorflow/core/common_runtime/device/device_event_mgr.cc:221] Unexpected Event status: 1

No idea how to fix it.

This happens randomly while training, sometimes an hour in, sometimes 20 min. However if I can get it to go for longer than 4/5 hours it seems to run fine.

My drivers are up to date, my cuda is up to date...

I am not the first person with this issue however I cannot find an actual solution
https://github.com/tensorflow/tensorflow/issues/46247
https://forum.faceswap.dev/viewtopic.php?t=2591


Re: Tensorflow error happens randomly while training

Posted: Tue Oct 17, 2023 11:29 pm
by bryanlyon

As the previous message said, this is a driver issue. Nothing to do with Faceswap and nothing we can do to prevent or work around it.

As far as I've been able to find on it though, it's almost 100% an overclocking issue. Chances are the card you have is factory overclocked and it's hitting instability because of it. Clock it back to Nvidia's recommendations and I think you'll see this message disappear completely.