* support sparse of adam, *test=kunlun

* add pre-commit-config.yaml

* support sparse of adam in KL2,*test=kunlun

* support sparse of adam in KL2, *test=kunlun

* modify xpu.cmake, *test=kunlun

* support sparse of adam, rm some wait, *test=kunlun

* support sparse of adam, rm some wait, *test=kunlun

* support sparse of adam, *test=kunlun

* support sparse of adam, *test=kunlun

* support sparse of adam, *test=kunlun

* support sparse of adam, *test=kunlun

* support sparse of adam, *test=kunlun

* support vs2019 compilation in windows

* not modify pow_op's original compute logic

Add exp, abs_grad, reciprocal, reciprocal_grad operator for XPU and update xpu2_op_list.h,test=kunlun (#38570)

* add argsort/scatter for kunlun

* update test_scatter

* update xpu.cmake

* update xpu.cmake

* fix scatter

* add infrt code

refined with Paddle's code style.

* rename CinnRtConfig to InfRtConfig

* rename CinnRt to InfRt of some code

* rename CINNRT to INFRT

* remove unnecessary code

* replace CINN to INFRT in the source code

* replace all "cinn" in code to "infrt"

* remove some const_cast

1. Modify git tag for CINN
2. Support compile option "-DWITH_CINN=ON, -DWITH_TESTING=OFF"

* 添加xpu指定url参数

* add expand_v2/expand_as_v2 for kunlun

* update expand_as_v2

* update expand_as_v2

* support float16/bool

* update xpu.cmake

* block xxhash warning of c4711

* modify according to zhouwei's comment

* fix syntax error

* make third_party's cmake get source code directly 2

* modify according to zhouwei's comment

* eager needs mkldnn to compile

* rename TensorBase interface data_type() to dtype()

* rename type to dtype of TensorMeta

* merge the code

* merge the code

* fix the problem when merge conflict

* fix bug of ci caused by type of tensor_meta

* changes cmake to clear cache

* upgrade oneDNN to v2.4-rc

* Removed failing test

* Revert "Removed failing test"

This reverts commit 60e70e717fac2c86b7beb24dfa1343a5804ea455.

* Remove most tests for debugging purposes

* Update hash to oneDNN 2.4

* Revert test change

* Update oneDNN to 2.4.2

* Update oneDNN to 2.4.3

* Change oneDNN version to 2.3 for Jenkins test

* Revert "Change oneDNN version to 2.3 for Jenkins test"

This reverts commit 0b176defc3b63f65dd0ba85873a018534f287000.

* Update oneDNN to 2.4.4

* Change version of oneDNN to 2.3 for new Jenkins test

* Revert "Change version of oneDNN to 2.3 for new Jenkins test"

This reverts commit e005a0f78f2b41cdcf4d7de3a21df7f910b78268.

* Add macro required by CINN.

* Remove CMAKE_BUILD_TYPE form cinn.cmake.

* Update the batch size used in test_resnet50_with_cinn.py.
* Enable more debug info.

* support different precision in kunlun

* minor

* minor

* minor

update cinn commit id tag to the newest one for fix some bugs

* Update the content of `test_parallel_executor_run_cinn.py`.

* Fix some bugs in the topological sort and `CreateNewSubGraph`.

* Update the CINN commit id used by Paddle.

* Update the unit test to `add+relu`.

* Update according to reviewers' suggestion.

* add some ops to train ssd on kunlun

* add some ops to train ssd on kunlun

* add some ops to train ssd on kunlun

* update cast op unittest

* update cast op unittest

* update cast op unittest

* update xpu cmake

* update cast unittest

This PR added some changes to match the CINN change for compilation. It also tried to fix JiangCheng's Problem in PR: https://github.com/PaddlePaddle/Paddle/pull/36100

These changes include:
1. Set `CINN_GIT_TAG` to a newer tag
2. CINN now just `make cinnapi -j`
3. We have to add `-DPY_VERSION=${PY_VERSION} -DWITH_TESTING=ON` to CINN cmake args
4. For CINN's third party dependencies, we could just include headers without target_link_libraries
5. Moved `cinn.cmake` from `paddle/cmake` to `paddle/cmake/external` to match old style. External folder contains `lite`, which is the same level of `cinn`
6. CINN added `-DNAMESPACE=cinn_gflags` in `gflags.cmake` to have different gflag namespaces between CINN and Paddle. It solved re-define problem.
7. Change namespace of `::google::` in gflags to `::GFLAGS_NAMESPACE`

Add Tokenizer related functionalities for Transformer model in order that the process of training and predicting is consistent.

* support the text string as an input Tensor
* support the "VOCAB"unordered_map<wstring, int> as an input Tensor to lookup tokens
* Tokenizer used for BERT. This tokenizer applies an end-to-end, text string to wordpiece tokenization.
* It first applies basic tokenization, followed by wordpiece tokenization.

* [NPU] update inference cmake, test=develop

* address review comments, test=develop

* fix compile error when WITH_ASCEND_CXX11 ON, test=develop

* support bert when exists matmul_v2

* update

* add argmax and iou_similarity for kunlun

* add argmax and iou_similarity for kunlun

* add argmax and iou_similarity for kunlun