* refine vgg proto && add report

* refine proto

* add proto for 2 machine

* add proto for 2 machine 4 GPU and measurement of reduce time for 2 machine 8 GPU

* refine protos

* remove comment

* out delay regst

* no delay edge to kMdSave task

* fix bug when elem cnt in model < split num

* add checker

* let opencv depend on libjpeg-turbo

* fix libjpeg-turbo dependency

* remove copy by itor

* use already existed copy method

* delete forward field override and use CopyField instead

* remove needless

* use libjpeg-turbo

* refine format

* parse /proc/meminfo

* CHECK for getMemInfo

* refine strToToken

* for review

* use onefloe cast

* set regst max=min when same stream

* move fix regst num to task node

* kCudaThreadsNumPerBlock can divided by temp_col_num

* refine type

* __syncthreads only sync threads in a block. we need to split data correctly

* Refactor random generator

* refactor rng

* remove random_generator_.reset if branch

* combine macro and remove const

* add declaration for generalized rng

* remove generalized rng definition

* move data tmp in reduce sum to buf

* change var

* change way to calc buf size in reduce sum

* A BnInOp2Blob wrapper is better than GetOutDiffBlob

* sv

* format