* add norm cpu

* update code;

* norm bug fix

* move norm op to pten; test=develop

* move norm op to pten; test=develop

* add norm util; test=develop

* fix norm npu bug; test=develop

* fix norm kernel bug; test=develop

* move kernel args to pten; test=develop

* move kernel args to pten sig; test=develop

* add more int id type support for embedding

* add ut

* add more ut

* fix ci error

* add fuse_relu_depthwise_conv_pass unittest

* fix atol and rtol

* fix according to review

* Add FuseOptimizerPass and fuse_adam_pass unittest

* add sgd and momentum unittest

* add fuse_optimizer_pass

* close amp

* close amp

* update

* fix run on two cards

* Update test_dist_fuse_adam_pass.py

* Update test_dist_fuse_momentum_pass.py

* Update test_dist_fuse_sgd_pass.py

* Create test_dist_fuse_sgd_pass.py

* Create test_dist_fuse_sgd_pass.py

* Create test_dist_fuse_sgd_pass.py

* Update test_dist_fuse_adam_pass.py

* Update test_dist_fuse_momentum_pass.py

* Update test_dist_fuse_sgd_pass.py

* Fix quantization next op findings

* Corrections according to the review

* - 38126 potential fix

* - fix

* - build fix

* - another candidate fix

* - compilation fix

* - another fix

* - Fix to activation of NHWC being first oneDNN op in chain on oneDNN ops

* - compilation fix

* - added NHWC reotating for elementwise being first op

* - compilation fix

* - compilation fix

* - Added UT

* - cosmetic fixes

* find gpu kernel in pten factory; test=develop

* check in functional kernel first; test=develop

* delete gloo connect retry

* the_one_ps dirs reconstruct

* .

* .

* create the_one_ps dirs

* create the_one_ps dirs

* create the_one_ps dirs

* create the_one_ps dirs

* create the_one_ps dirs

* create the_one_ps dirs

* the one ps dirs modify

* the one ps dirs modify

* the one ps dirs modify

* the one ps dirs modify

* refactor ps optimize

* refactor ps optimize

* refactor ps optimize

* .

* .

* .

* .

* .

* .

* refactor theoneps

* the_one_ps

* add ps pass unittest

* add ps pass unittest

* ps unitest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* ps unittest frame

* add cpu_async_ps_mode test

* add cpu_async_ps_mode test

* add cpu_async_ps_mode test

* ps unittest ready

* ps unittest ready

* solve dist_pass init conflict

* solve import CommContext error

* unittest ok

* implement AllocateFrom

* solve setup.py.in conflict

* solve conflict

* solve conflict

* solve conflict

* .

* .
Co-authored-by: Nzkh2016 <zhangkaihuo@baidu.com>

* add concat & split

* add concat kernel

* add concat unittest

* add split unittest

* gpu_context..

* update

* update

* update

add downpour_ctr_accessor

* Replace clip, bce_loss, full and full_like with elementwise

* adapt selectedrows in execution

* impl selected rows branch

* support selectedrow in infershape utils

* fix device compile failed

* fix new exe test failed

* revert some changes

* Rough implementation for experiment

* Support allocate cuda managed memory

* Fix CI error

* Modify UT

* Check whether support memory oversubscription

* Fix ROCM Compile error

* Fix ROCM Compile error

* Fix UT cuda_managed_memory_test

* Set UT timeout to 40

* Add UT OOMExceptionTest

* Set UT timeout to 50

* hack custom op

* add ut

* skip windows ci

* Added adam kernel

* CI rerun

* refactor custom op kernel func and utils

* add output sync

* adapte tensor* in utils

* fix windows symbol error

* remove kernel alias name

* fix depreacted error

* fix deprecated failed

* fix mean error

* resolve conflict

* fix windows failed

* geo depends

* add memory geo table

* fix

* change unary infermeta

* change other infermeta

* change all infermeta format

* resolve conflit

* fix test failed

* resolve reshape conflit

* fix compile failed

* adapt auto api gen

* fix reshape failed

* fix concat failed

* resolve conflict

* upgrade _get_all_register_op_kernels

* add ut

* support xpu/npu

* fix device id

* enhance TransToFluidPlace

* fix compile

* Add fp16 support for scale/bias for fused_layernnorm_residual_dropout_bias op.

* Remove useless code.

* Remove useless code.

* Optimize layer_norm fwd when cols is 1024.

* Remove useless code.

* Minors.

* Minors.

* Modifications accordding to reviews.

* Minors.

* Optimize layer_norm bwd kernel when cols is 1024.

* Polish layer_norm_bwd_1024 kernel.

* Limit ln_bwd_1024_kernel to paddle_with_cuda.

* Fix double type compile error.

* Add optimization of ln bwd for fused_dropout_add_ln op.

* Polish codes.

* Add XPU compiler for paddle, test=develop

* clean code

* clean useless code

* clean useless code

* clean useless code

* test

* add include path

* use clang compiler

* xpu2.cmake

* XPU2 compiler passed

* update

* update after pten

* combination the WITH_XPU and WITH_XPU2

* update the fuse operation in WITH_XPU and WITH_XPU2

* update

* update

* update

* fix the merge error

* update

* update the code

* update the code

* add run_kp_kernel flag

* update

* update

* fix prepared type_ bug

* clean and update the code

* reset the kernel_primitives

* update

* clean the code

* delete useless comment

* fix the bug in WITH_XPU

* update

* update

* modify the abi

* delete some useless code

* Parameter automation in xpu compilation

* Parameter automation in xpu compilation

* delete kps in cmake

* delete useless comment

* clean the code

* clean the code