i55code opened a new issue #17678: Cuda10.2 installation error URL: https://github.com/apache/incubator-mxnet/issues/17678 Hi, I have followed this post (https://github.com/apache/incubator-mxnet/issues/17058) to install the nightly convenience builds using: pip install --pre mxnet -f https://dist.mxnet.io/python/cu102mkl However, when I test the following code: >>> import mxnet as mx >>> a = mx.nd.ones((2, 3), mx.gpu()) The error message is : >>> import mxnet as mx >>> a = mx.nd.ones((2, 3), mx.gpu()) [20:56:28] src/imperative/./imperative_utils.h:92: GPU support is disabled. Compile MXNet with USE_CUDA=1 to enable GPU support. terminate called after throwing an instance of 'dmlc::Error' what(): [20:56:28] src/imperative/imperative.cc:81: Operator _ones is not implemented for GPU. Stack trace: [bt] (0) /home/.local/lib/python3.6/site-packages/mxnet/libmxnet.so(+0x307d3b) [0x7f1a6fb0fd3b] [bt] (1) /home/.local/lib/python3.6/site-packages/mxnet/libmxnet.so(mxnet::Imperative::InvokeOp(mxnet::Context const&, nnvm::NodeAttrs const&, std::vector<mxnet::NDArray*, std::allocator<mxnet::NDArray*> > const&, std::vector<mxnet::NDArray*, std::allocator<mxnet::NDArray*> > const&, std::vector<mxnet::OpReqType, std::allocator<mxnet::OpReqType> > const&, mxnet::DispatchMode, mxnet::OpStatePtr)+0x6bb) [0x7f1a72cb1c5b] Aborted (core dumped) Please let me know how to install mxnet properly for Cuda 10.2 in ubuntu 18.04. Thanks!
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services
