Fix problem that target name already exists when there isn't model data cache, test=develop (#27142)

* Modify data download function, and support unittests of inference APIs on windows, test=develop

* The import error compatible with py2 and py3, and fix unittests problems of inference APIs on Windows, test=develop

* Add mkldnn bfloat16 option to C-API

* Add test for bfloat16 gpu

* Change coverage test

* Repair capi_gpu test

test=develop

* add stack_op to CMakeLists

* add dim=3 support for scale op

* add trt stack op, test=develop

* remove debug message

* add stack plugin serialize

* remove slice, scale op, will add later

* enhence error message

* revise trt ernie test to conver the stack op CI testi, test=develop

* add stack op serialization

* fix test shape after adding stack op

* remove slice op, will add after implementing serialization

* roll back to min_graph=5 to avoid using slice op

* fix scale op output layer

* implement stack op createPlugin

* use workspace and move the defination to .cu

* move stack plugin creator definition to .cu, test=develop

This reverts commit 02083bda.

* support trt dynamic shape int8

* add unittest

* add support for sigmoid; adapt to trt6+ api

* refine paddle inference api
Co-authored-by: Nnhzlx <nhzlx.dragon@gmail.com>

* Add mkldnn bfloat16 option to C-API

* Add test for bfloat16 gpu

* Change coverage test

* Add the option to execute unit tests only at night

* set ut nightly label for 3 cases.

* Fix the cmake-function named inference_download_and_uncompress with Windows, test=develop

* Fix some problems when remove limit of unittests on Windows, test=develop

* Using URL to download file instead of DOWNLOAD_COMMAND. test=develop

test=develop

* Add laxical GRU unit test

performance works

* Get model accuracy

* model and data name to be confirmed
test=develop

* update model name and output format
test=develop

* update according to reviews
test=develop

* add accuracy check

* accuracy check between native and analysis
test=develop

* fix a reading bug, fix gru passes sequence
test=develop

* fix passes sequence
test=develop

* Change use_quantizer attribute name and data type

* Fix problem with setting attribute

* Add changes due to review

* Small change in function

* Restore use_quantizer attr for compatibility

* fix trt plugin registry without trt lib

* support trt4

* refine code style

* develop dynamic shape serilization

* add test param for gelu

* fix bugs

* delete redundant comments

* debug

* fix conflict. test=develop

* fix bug. test=develop

* add trt dynamic shape serialized support

* fix ernie serialized bug
test=develop

* fix codestyle
test=develop

* fix bug
test=develop

* fix bug.test=develop

* modify cmakelist test=develop

* fix bug
test=develop

* fix error message.  test=develop

* fix trt register plugin based on pr#25003

* add trt dynload

* fix deserialization bug of not finding plugin registration

* refine code style

* recover engine key in tensorrt_subgraph_pass

* for ci coverage

* add unittest for deserialization
Co-authored-by: Nhaozech <chenhaoze94@gmail.com>

* fix windows test
test=develop

* fix ci
test=develop

* update ResetHolder, test=develop

* add TensorShare for lite engine, test=develop

* tensor data changed from copying to sharing, test=develop

* supports xpu runtime, test=develop

* fix code styles, test=develop

* fix windows ci
test=develop

* fix ci error

* add api to clear intemediate tensors in analysis predictor. test=develop

* add python api. test=develop

Use vector instruction (LDG.128) to improve qkv transpose. It
provides 1.4X speedup at same GPU base frequency.
test=develop