| The AveragedLoss op takes a single
| 1-D input tensor input and returns
| a single output float value output.
| The output represents the average of
| the values in input. This op is commonly
| used for averaging losses, hence the
| name, however it does not exclusively
| operate on losses.
|
| Github Links:
|
| - https://github.com/caffe2/caffe2/blob/master/caffe2/operators/loss_op.h
|
| - https://github.com/caffe2/caffe2/blob/master/caffe2/operators/loss_op.cc
|
| AveragedLoss takes in the input and
| produces the output loss value as the
| average of the input.
|