CompactBilinearPooling-Pytorch icon indicating copy to clipboard operation
CompactBilinearPooling-Pytorch copied to clipboard
verified publisher icon DeepInsight-PCALab

multi GPUs error: RuntimeError: arguments are located on different GPUs

Open JingyunLiang opened this issue 7 years ago • 3 comments

For multi GPU, it outputs:

RuntimeError: arguments are located on different GPUs at /pytorch/torch/lib/THC/generic/THCTensorMathBlas.cu:236

How to fix it?

JingyunLiang avatar May 21 '18 05:05 JingyunLiang

Currently I don't have a multi-gpu environment, sorry about that. You are welcomed to create a PR if you have a solution.

Just by any chance it works, try reduce the batch/tensor size.

dichen-cd avatar Jun 12 '18 08:06 dichen-cd

Hello, is your problem solved? Can you train on multiple GPUs? Thanks

mxtx0509 avatar Jul 12 '18 08:07 mxtx0509

I encountered the same problem during multi-GPU training.

mxtx0509 avatar Jul 12 '18 08:07 mxtx0509