Please explain the issue:
" mxnet.base.MXNetError: [12:56:29] src/operator/nn/./…/linalg_impl.h:166: FP16 gemm on cpu not implemented!"
Hi,
the gemm operator is not implemented yet for this CPU x FP16 data.
I assume you are trying to use (automatic) mixed precision? If so, you should run the failing action on a GPU instead of a CPU.
hth,
Lieven