8 Commits (dace68ac6d95e6dcffecd0edca5cd5ae290bba2f)

Author	SHA1	Message	Date
qingqing01	32d881beab	Optimize the rowwise add function.	9 years ago
qiaolongfei	682eee40cb	fix math_function warning	9 years ago
Yu Yang	7e214b4985	Speed up ColwiseSum in CPU (#6834) ... * Remove unnecessary reshape in ColwiseSum Speed up 12s -> 10s. * Hand write ColwiseAdd in CPU	9 years ago
QI JUN	61ec0b9516	Refine device context (#6433) ... There are mainly following fixes: - take `DeviceContext` as the template parameter of math functors and OpKernel instead of `Place` - remove `eigen_device` interface in base class `DeviceContext` - remove `GetEigenDevice` interface in `ExecutionContext` and base class `DeviceContext` - remove unused `platform::EigenDeviceConverter` - rename `REGISTER_OP_GPU_KERNEL` to `REGISTER_OP_CUDA_KERNEL` - rename `USE_GPU_ONLY_OP` to `USE_CUDA_ONLY_OP`	9 years ago
dangqingqing	2673657684	Move RowwiseAdd functor to math_funcion and Add ColwiseSum functor.	9 years ago
dangqingqing	91d4fc6941	Fix compling for softmax_with_cross_entropy_op.	9 years ago
dangqingqing	5f21709938	Fix bug.	9 years ago
dangqingqing	f5e367655e	Use G++ to compile some cu operators.	9 years ago