I'm a newcomer, arranged to maintain a system using this py-R-FCN. I want to use this system on a machine with multi-GPU.
But I see in the Read-me, it suggests to use mxNet for multi-GPU train/test.
Now the new version caffe supports multi-GPU, does py-R-FCN run on multi-GPU automatically?
Hi,
I'm a newcomer, arranged to maintain a system using this py-R-FCN. I want to use this system on a machine with multi-GPU. But I see in the Read-me, it suggests to use mxNet for multi-GPU train/test. Now the new version caffe supports multi-GPU, does py-R-FCN run on multi-GPU automatically?
Thanks!