Implementation of batch norm

I am looking for where batch norm is computed in the code on the pytorch github, but I cannot find it anywhere. Could anybody link me to the file its implemented?

@bananacode Are you looking for the kernel implementation of batch_norm, or are you looking for the Python / C++ API entry point of batch_norm?

cuDNN bindings are at https://github.com/pytorch/pytorch/blob/700109eb630b79fd65cb93becb7f2d14f93bdb5c/aten/src/ATen/native/cudnn/BatchNorm.cpp

Sorry I should’ve specified. I am looking for the CPU implementation of batch_norm.