* save infer model bug fix

* test=release/1.3

* Impl fp16 compute kernel for slice_op

* Use data() to replace mutable_data()

test=release/1.3

test=release/1.3

test=release/1.3

* fix params with only 1 dim
* test=develop

* fix group_norm (#15727)

This reverts commit 08a8c1d1.

* test=develop, change md5 for patchELF

* test=develop, update protobuf

* test=develop, update protobuf in Dockerfile used by CI

* test=develop, install requirements before start for Linux

* test=develop, uninstall protobuf on linux brefore install latest version of it

* test=develop, fix pip

* test=develop, install protobuf in linux

* test=develop, fix patch ELF install failed

* test=develop, change md5 for patchELF

* test=release/1.3, cherry-pick

test=release/1.3

test=release/1.3

test=release/1.3

1.Remove the code for setting mkldnn environment in the test_calibration.py;
2.Update the cmake file for MKLDNN environment enabling;
3.Update the INT8 inference doc.

test=release/1.3

test=release/1.3

test=develop

test=release/1.3

test=release/1.3

test=develop

* expose peak gpu memory API to python test=develop

* add unittest for peak gpu memory monitoring test=develop

* add pybind change test=develop

* add mutex to gpu mem usage monitor test=develop

* update benchmark flag definition file test=develop

* tweak unittest for memory monitoring test=develop

test=develop

* Add calibration README; test=develop

* Add INT8 calibration support in Paddle package; test=develop

* Enable pool2d operator for a ngraph engine
test=develop

* Update
test=develop

* move var strusted to vars_distributed.py, add optimizer's block name, test=develop

* rename optimzier's seems complex, revert it, test=develop

* replace * with details, test=develop