Details of implementation

Where can I get the detailed implementation of linear regression in mxnet. I have looked at regression_output.cu, regression_output-inl.h , regression_output.cc. But it uses namespace mshadow and I cannot get into the details of how the gradient calculations and forward passes are calculated. Kindly consider helping me in this regards. Thanks in advance for your help.

consider
in_data[reg_enum::kLabel].Size()/in_data[reg_enum::kLabel].shape_[0];
Tensor<xpu, 2> out = out_data[reg_enum::kOut].FlatTo2D<xpu, real_t>(s);
Tensor<xpu, 2> grad = in_grad[reg_enum::kData].FlatTo2D<xpu, real_t>(s);
Tensor<xpu, 2> label = in_data[reg_enum::kLabel]

inside regression_output-inl.h, where are these functions defined ?