OMP_NUM_THREADS error.

[saikrishnadas]

I get an error while trying to use distributed training. I have 4 GPUs(Tesla T4) and error shows when using a p7 model. Tried switching to single GPU and same error occurs. But it works with csp model with one gpu.

Error log : **Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed.**

[WongKinYiu]

this is not error message, all of ddp training will show this default message.

[saikrishnadas]

But my training process is exited after showing this.

[saikrishnadas]

Do you have a solution for this? This usually happens when I use more than one GPU with p7 model @WongKinYiu

[RohitKeshari]

I also encounter with this problem, @saikrishnadas did you got any solution for this?

[marcelampc]

My training isn't even killed. It just freezes before it begins.