Hi, there. Does mxnet have a command line option for enabling GPU?
In Tensorflow, I can use export CUDA_VISIBLE_DEVICES
. In DyNet, I can use --dynet-gpus
cli option.
In mxnet, do I have to implement such switches by myself?
Thanks.
Hi, there. Does mxnet have a command line option for enabling GPU?
In Tensorflow, I can use export CUDA_VISIBLE_DEVICES
. In DyNet, I can use --dynet-gpus
cli option.
In mxnet, do I have to implement such switches by myself?
Thanks.
Wow, that’s convenient. I mean enabling, sometimes I develop models on mac and run it on server, so it’s convenient to active GPU by default.