Train_network stops working after ~5000-7000 iterations

I’ve successfully labeled points and now when I try to train them on Jupyter Notebook I get a timeout error anywhere between 5000 and 7000 iterations and the kernel stops. I’m using the dlc-windowsCPU environment. Does anyone know of a way to fix this? I’m feeling like it’s due to Jupyter Notebook but I’m not sure what to try. Thank you!

are you sure you are inside the dlc env when you run the notebook?

Top left corner should have:
image

i.e. go to Kernel > Change Kernel

I’m still getting this problem after reinstalling the environment. Do you know of any other reasons that could be causing this? Thank you!

If your computer goes into power save mode, or sleep, etc, that can happen… but otherwise, I am not sure. Can you post the output of what happens at 5-7k iterations?

It happened to be due to overuse of the server I’m training on. I logged everyone else off and reset it and it worked perfectly. Thank you for your help!

1 Like