test=develop

test=develop

* add launch mp distributed mode module test=develop

* delete unused file test=develop

* refine usage test=develop

* refine usage test=develop

* move distributed package test=develop

* add to whl package test=develop

refactor optimize pass.

add device info to tensor

test=develop

add elementwise_xxx_grad for inplace optimize

* fix compile issue with_distribute test=develop

* simplified logic test=develop

* use ngraph dependency test=develop

* set cpu only test=develop

* update test and eliminate fp16 test test=develop

fix anakin compile dependency

Fix Pr #15296 async buffered reader

test=develop

test=develop

test=develop

speed up box_coder in CPU

fix ctr reader read svm data

fix the lib_any dependency

test=develop

test=develop

fix ci broken randomly and disable some warnings

fix the build issue on gpu mode for win

test=develop

* Added accuracy ngraph op test=develop

* fixed name type test=develop

* add skip send.recv test=develop

* enhanced print message. test=develop

* rerun ci. test=develop

test=develop

size_t is an unsigned integer, with a conversion rank
 larger than int, therefore in the following expression
 the int value was promoted to size_t, making it a
 subtraction of unsigned values. The result of such
 a subtraction is also an unsigned value.

Revert "Async double buffered py reader"

test=develop