* implement MHA order same as training

* fix fp16 compile issue on old architecture
Co-authored-by: Nzlsh80826 <rewang@nvidia.com>

* fix airank bert emb order

* move input num check to converter

* add input num check

* add unused var check white list

* add trt runtime version check

* use different wrap, and change to major version check

* fix tc trt shape

* fix fc dynamic shape

* add fc shape assert

* update

* move semantic checks to op_teller

* more ops

* more ops

* revert block related change

* part1

* revert activation

* remove if

* remove const_cast

* reslove conflict

* remove const_cast

* delete useless var

* replace vlog(1) with vlog(3), replace assert with PADDLE_ENFORCE

* down to 19 files

* support ernie trt-int8 for inference

* fix reshape

* extend multiclass_nms unittest timeout threshold

* adjust timeout to 200s

* temporarily disable multiclass_nms trt op teller

* fix yolobox teller condition

* fix cuda double free bug

* update trt engine addplugin name.

* update

* add anchor generator op plugin

* add anchor generator unit_test

* remove dbg info

* remove redundant line

* replace assertion with paddle enforce

* dynamic plugin replaces assertion with paddle enforce

* anchor generator support dynamic shape on spatial axis

* anchor generator test with fp16, dynamic shape

* add anchor generator test all

* add back main

* reduce test input size to not exceed the timelimit of ci

* change super to InferencePassTest for python2 compatibility

* reuse paddle operator anchor generator

* move creator construct to header with default

* add cuda ifdef

* reduce line

* change super to InferencePassTest for python2 compatibility

* fix anchor generator fp16 serialize setting

* split unittest from test_all

* restrict anchor generator input format before version 7234

* anchor generator only support greater than trt7.1

* change min_graph_size to 2

* min_graph size to 3 if dynamic shape

* reduce dynamic shape size to avoid trt search tactic too long to exceed time limit

* remove anchor from fetch list

* anchor generator support all trt version

* fix memory not allocated but if serialized

* fix batchnorm when inpu dims < 3

* add unittest for batchnorm dims = 2

* yolobox converter and plugin

* yolobox unittest

* add dynamic shape restriction

* fix git merge log

* add roi_align_plugin

* add roi align unit_test

* add roi align serialization

* remove roi align static plugin because of batch dim issue

* refine roi align unittest and add fp16/serialization

* add trt roi align condition to op_teller

* refine error message

* remove unnecessary reshape layer

* trt affine channel converter

* add trt affine channel base test

* add trt affine channel NHWC

* remove asterisk for python2 compatibility

* trt affine channel converter

* add trt affine channel base test

* add trt affine channel NHWC

* remove asterisk for python2 compatibility

* fix rebase

* move LodTensor to Tensor

* add dbg info

* affine channel converter only support NCHW

* scale,bias are parameters, use create_parameters api

* reduce test input size to not exceed the timelimit of ci

* refine affine channel unittest and add serialization/dynamic test

* change super to InferencePassTest for python2 compatibility

* change super to InferencePassTest for python2 compatibility

* fix affine channel fp16 serialize setting

* add multiclass_nms

* add multiclass_nms unittest

* add default enable_tensorrt_oss option

* refine multiclas nms unittest and add serialization/dynamic test

* change super to InferencePassTest for python2 compatibility

* refine multiclass nms unittest

* move out dynamic shape test due to ci timelimit

* fix tensorrt output varible reshape

* move padding shape x 1 x 1 in ernie to qkv and fc

* update layer name

* fix softmax when input is dynamic, fc not padding any more

* fix varlen

* move fc x_dim assert to op_teller

* nearest_interp op converter w/ dynamic/static

* fix data_layout include

* add trt nearest unit_test

* add nearest_interp NHWC test

* update trt nearest interp nhwc testcase

* remove asterisk for python2 compatibility

* add empty line to prevent conflict

* nearest_interp op converter w/ dynamic/static

* fix data_layout include

* add trt nearest unit_test

* add nearest_interp NHWC test

* update trt nearest interp nhwc testcase

* remove asterisk for python2 compatibility

* add empty line to prevent conflict

* change the priority of out_h, out_w

* trt gather converter

* add trt gather unit_test

* support batch axis concatenation when using dynamic shape

* opteller can't return true early, or some test will not be executed

* add serialize unittest

* fix element_op trt plugin serialize bug

* add n-d input support for trt scale converter

* add flatten for ut

* fix dims

* add group norm plugin

* fix compile problems

* move concat axis check to trt op teller

* add nbDims for scale and bias nv dims

* add group norm unit test

* fix unittest

* add trt version restriction for group norm op teller

* fix unittest

* update trt int8 calibrator to IEntropyCalibratorV2

* add delele opt_cache for trt_split_converter_test

* fix split trt plugin initialize

* update

* use iwyu clean include second time, test=develop

* fix trt plugin clone and initialize bugs

* fix unit test error

* enable trt in ci py3

* update unittest timeout