* add inplace op adaptation

* optimize inplace logic and fix bugs when run kernel that has args of vector<DenseTensor>

* refactor logic that transform variable to densetensor

* update func name

* [NPU] add int64 support for scatter op

* [NPU] delete debug codes

* [NPU] optimize codes

* first commit

* more uts

* file name duplicated

* timeout

* Update CMakeLists.txt

change TIMEOUT from 120 to 240

* Update CMakeLists.txt

more time

* Update CMakeLists.txt

timeout

* Update CMakeLists.txt

60s

* Handled dispensable tensors in AutoCodeGen for Eager Dygraph

* Enabled Eager Dygraph AutoCodeGen for 500+ existing ops

* fix inplace addto pass

* update

* fix ut

* improve ci coverage

* fix musl ci compile error

* optimizer __call__ to make dygraph faster

* fix return type

* remove cpp Layer

* add align for WorkQueue

* add spinlock

* merge develop

* merge

* Add EventsWaiter

* Revert "Add EventsWaiter"

This reverts commit e206173aa9be7401b83a53581627bfaf557c8fb2.

* update HostEventTracer

* update HostEventTracer

* fix c++17

* update

* update

* update

* update

* fix bug
Co-authored-by: Nliutiexing <liutiexing@google.com>

* add prior_box for kunlun

* update

* update CMakeLists

* fix fc_fuse

* modify cmake notest,test=windows_ci

* retrigger all the ci

* add angle_op

Modify ShareTensorWithCinnBuffer by callback to save memory

* refactoring matmul hierarchy

* review fix

* review fix

* review_FIX-part2

* Add new unittests

* Replace I with O channel for filter groups

* Undo changes affecting other operators

* Fix oneDNN namespace typo

* Fix code format error

* add regularation and Nesterov for mergerd_momentum

* refine unittest for use_nesterov attr

* refine op check

* refine code

* fix bug

* refine code of regularization_flag

* delete useless code

* Refactored eager legacy namespace

* Fixed namespace issues

* Enabled performance benchmark tests for Eager Dygraph

* Protected CUDA tests with macro

* Fixed dependency issues for windows-ci

* support data_format='NHWC' for prelu channel mode

* add scope_guard

* 1. fix control flow cases 2. fix calc_gradient

* add pten reduce kernel

* add reduce_sum kernel

* update attribute args and order

* make out dtype undefined

* fix empty input error

* merge develop branch

* rename sum as reduce function

* rename sum as reduce function

* fix reducekernelImpl args error

* add reduce cuda kernel

* modify dims type to const &

* remove unsed log

* fix reduce_all out eigen function error

* remove unused codes

* add the missing sum api define and testcase

* merge develop branch

* fix sum test axis value error

* replace pten mean kernel with reduce_mean

* revcover meam cuda to original implement

* add expand_v2/expand_as_v2 for kunlun

* update expand_as_v2

* update expand_as_v2

* support float16/bool

* update xpu.cmake