test=develop

* clip softmax for fp16

* updates

* fuse xent support fp16 test=develop

* wip

* wip

* add simple row reduce

* wip fp16 accurate softmax

* add accurate softmax kernel for fp16 test=develop

* update test=develop

* fix cpu build test=develop

* update api.spec test=develop

* follow comments test=develop

* fix build test=develop

* fix trt build test=develop

* fix inference build test=develop

* fix merge test=develop

* update test=develop

* try fix build test=develop

* fix build test=develop

* rename real_exp test=develop

* fortest

* remove hacky kernels test=develop

* clean up test=develop

Set correct TBB library name in debug build 

reduce threads number to avoid analyzer_rnn1_tester hang in CI

fix demo ci bug

Enforce dim check in data feeder

fix dist sparse l2 decay

Set correct TBB library name in debug build and remove warning related to rpath dependency from symlink.
test=develop

Fix ngraph cmake warning 

Fuse/seqpool concat 2

1. trt_demo bug
2. trigger exit when exists a bug

test=develop

fix analyzer_test runs error in native_config

test=develop

try unify Executor and ParallelExecutor

test=develop

test=develop

test=develop

use the new MKLDNN repo url

test=develop

test=develop

refine seqpool op

test=develop

Since conv_fusion_op is not exposed into Python, remote the env flag in __init__.py
test=develop

Remove op handle lock and fix var

Fix same name func framework::ToTypeIndex

fix debug build error

PADDLE_WITH_NGRAPH was removed from the code

test=develop

test=develop

fix analysis_tester bug

test=develop

test=develop