* fix trt plugin clone and initialize bugs

* fix unit test error

* enable trt in ci py3

* update unittest timeout

* Add cache for Transformer encoder.

* Bug fixed.

* add unittests for transformer encoder.

* update gather_tree doc, test=document_fix

* update sample code, test=document_fix

* remove tensor type, test=document_fix

* upgrade gather_tree to core.ops

* update gather_tree unittests

* added external reorder to profiler

* resolved conflict

* added enable_static

* initial version of lstm, not working yet

* added lstm to operators.cmake

* added vanilla lstm mkldnn op

* added peephole weights integration

* minor changes

* added formatting

* added fusion_lstm_mkldnn to static_whitelist

* added formatting

* removed comment

* moved use_peepholes attribute inside is_cached block

* reverted wrong changes

* minor formatting change

* minor changes

* changed stream handling

* minor change

* added datatype to GetExpectedKernelType()

* added reading stream from TLS

This reverts commit d834f4e6.

* added external reorder to profiler

* resolved conflict

* added enable_static

* initial version of lstm, not working yet

* added lstm to operators.cmake

* added vanilla lstm mkldnn op

* added peephole weights integration

* minor changes

* added formatting

* added fusion_lstm_mkldnn to static_whitelist

* added formatting

* removed comment

* moved use_peepholes attribute inside is_cached block

* reverted wrong changes

* minor formatting change

* minor changes

* polish printing dtype

* fix special case

post quantize support insert fake_quantize_dequantize node before the OPs that will be used in VIS's faceid models (#30659)

test=develop

* add abs test case

* use std::abs to fix abs bug

* fix the abs bug

* fix abs bug

* Fix the bug in fleet amp_init.

* Fix the amp_init unit test.

* Fix the error of save_quantized_model

* support reduce_max op on kunlun

* support reduce_max op on kunlun

* support reduce_max op on kunlun

* support reduce_max op on kunlun

* delay the 'timeout' of 'test_static_save_load'.

* delay the 'timeout' of 'test_static_save_load'.

* rewrite abs op

* rewrite abs op and remove abs in activation

* remove abs register in old codes

* fix abs_grad type error

* fix abs double_grad output name error

* modify abs_grad, abs_grad_grad functor for windows building

* format code style

* fix the bug of result is nan when the divisor is zero

* add missing abs attr and add abs for float16

* add fleet amp.init()

* add unittest for fleet_amp_init

* delete empty line of pybing.cc, test=develop

* fix layers train eval bug, test=develop

* update, test=develop

* add paddle.

* add unittest

* add rmsprop_op_xpu test=kunlun

* modified rmsprop_op_xpu error code. test=kunlun

* support layer_norm fp16 in dygraph amp

* add ut

* refine code

* add 2.0 API: device_guard