* add DistributedFusedLamb op

* polish code

* fix compile error

* compatible with pten changement

* fix rocm compile error

* improve converage

* update upstream/develop

* fix cast_with_ptr.h

* add FLAGS_distributed_lamb_divide_nranks_when_allreduce=1

* fix clip before allreduce

* add use_master_param_norm

* code polish

* fix bug

* fix ROCM ci

Move paddle/fluid/operators/math/algorithm.h to paddle/pten/kernels/funcs and rename all references to symbols in it.

* #1 migrate dist-related type()-> dtype()

* move datatype function from pten -> fluid/framework

* change type() in imperative into convert(dtype())

* modify xx_tensor->type into xx_tensor->dtype

* change the set_type interface and the caller

* modify xx_tensor.type into xx_tensor.dtype

* fix mutable_data(place, dtype())

* change caller of mutable_data in pten and distributed

* change the caller of mutable_data in fluid/framework

* change the caller of mutable_data in imperative directory

* mutable_data: inference

* update the call of mutable_data

* transfer MakePenScalarArray MakePtenScalar ResetHolderWithType

* pass the compile. the next step is remove VarType in Pten

* fix all and remove VarType from pten. success in linux. Next task is other platform

* fix conflict with develop

* fix compiled error

* Fix reset conversion

* fix conflict

* fix compiled problem

* fix typo

* Fix << in tensor_utils.cc

* fix type->dtype

* fix unittest

* fix tensor init constructor

* fix DataTypeSize for BFloat16

* fix code style

* fix npu compiled error

* fix npu

* compile npu sucessfully

* fix conflict

* fix conflict
Co-authored-by: Nxiongkun <xiongkun03@baidu.com>

* move operators/math/math_function_* to pten/kernels/func
* namespace from `paddle::operators::math` to `pten::funcs`

* merge legacy to fluid

* Remove legacy code

* Remove legacy code

* Remove DataType test

* Using Tensor directly instead of using EagerTensor

* support gradient_accumulation

* make test_imperative_lod_tensor_to_selected_rows longer

* make test_imperative_lod_tensor_to_selected_rows longer

* Added adam kernel

* CI rerun

* Added selected_rows and rw_lock to pten

* Renamed the unit test target to fix CI

* Removed Class SelectedRows in Fluid, changed include/cmake relationship, use pten::SelectedRows in Fluid

* Remove rw_lock.h,rw_lock_test.cc in fluid

* Use pten::RWLock and pten::AutoRDLock, fix CI

* Use pten::SelectedRows

* Use pten::SelectedRows

* Fix to pass NPU CI

* Use pten::SelectedRows, to pass NPU CI

* To fix NPU CI

* To fix NPU CI again

* migration of functors in paddle/fluid/operators/eigen and paddle/fluid/platform/eigen_ext.h
* update path of data types like float16.h in includes in extensions.h

* 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

* fix mp

* support merged_momentum for mp

* Merged LoDTensor with Tensor,test=allcases

* Patched python level LoDTensor

* Patched python level LoDTensor

* Merge Tensor into DenseTensor

* Fixed namespace issues,test=allcases

* Fixed merge issues

* Fixed inference issues

* Fixed NPU test issues

* Fixed merge issues

* Added shared_ptr<Allocation> member & corresponding interfaces to Storage

* Removed original pten::Allocation from Storage and adjusted the interfaces accordingly

* Fixed issues with storage offset

* Used place to malloc allocation for TensorStorage

* [Unify Tensors PR #3]Ported framework::Tensor interfaces to pten::DenseTensor

* Fixed issues with place

* Added comments

* Moved mutable_data with stream argument to DenseTensor

* Added set_offset interface

* Fixed CI issues,test=allcases

* [Unify Tensors PR #4] Port LoDTensor interfaces to DenseTensor

* Removed friend class EigenTensor/EigenMatrix/EigenVector from Tensor

* Modified framework::Tensor to inherit from DenseTensor

* Reverted changes too pten_layout() interface

* Removed friend classes

* Rearranged cfunction calls from tensor.data<void>() to tensor.data()

* Fixed CI issues

* Fixed lite issues

* Fixed data() interface issues,test=allcases

* Resolved IsInitialized() issues

* Fixed ResetHolder() issues

* Fixed MKLDNN & Storage issues

* Resolved ShareBufferWith() issues

* Fixed LoD issues

* add multi tensor for adam

* add merged_adam op

* refine code

* refine adam compute logic

* support multi precision update for LAMB

* hide some api

* fix ci uts

* fix lamb output of dygraph

* remove some changes to some PR

* try to fix Py3 CI compile error

* fix test_imperative_optimizer, add lars ut, add layer_norm ut

* fix ut, fix format

* fix ut

* fix windows ci

* refine structure for cuda and rocm

* update

* update

* update

* update

* 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

* [NPU] reorganization for device API abstraction

* [NPU] delete old files

* [NPU] fix npu_collective_helper

* [NPU] fix collective_helper

* [NPU] fix ut

* [NPU] mod memory allocation and hccl_helper

* [NPU] fix place_type

* [NPU] split enfoce.h

* move acl* call into npu_info

* merge conflict

* fix merge

* merge conflict

* merge conflict

* copy beta pow to same place when skip_update=1

* fix xpu

* add pow2_warmup op

* remove contrib __all__

* add AttrT

* rename

* follow comments

* fix duplicate PADDLE_RESTRICT

* remove wrong restrict

* remove master_param_out __restrict__

* update

* merge momentum ops

* update

* add ut to improve coverage

* remove optimizer change

* fix error msg

* update ut

* add __restrict__ for CUDA

* update ut

* move merged_momentum_op to optimizer dir

* fix coverage

* A leap of try for cudaLaunchCooperativeKernel

* fix bugs

* Totally replace the lar cuda kernel

* Fix bugs

* a test for lars merge

* Adding las_op_momentum infer_shape

* Fix codes

* use avg_numel instead of max_numel to acquire grid num

* modify unittest files about lars op

* Finally converge when merged-lars works

* fix ctest files

* add merged_operation kernel when cuda version is older than 11

* Fix code style

* fix ctest failure

* fix error

* fix all ctest error and change lars compute code of cpu

* fix bugs on v100.

* revert python modififation about lars

* revert python modification codes

* A leap of try for cudaLaunchCooperativeKernel

* fix bugs

* Totally replace the lar cuda kernel

* Fix bugs

* fix code according to comments

* fix codes according to  review comments

* adding some function overload

* relocate the power operation.

* Create stateful OneDNNAXPYHandler object.

This makes it possible to call it multiple times without recreating the
oneDNN primitives every time.

* Prepare SGDOpKernel to reuse its implementation from OneDNN kernel.

* OneDNN SGD kernel.

* Update call to use new OneDNNAXPYHandler object api.

* Setup seed in proper place.

* Enable OneDNN kernel only for single case.

* For dense param and sparse grad.

* Small refactor.

* Enable oneDNN by op attr or by cmd line flag.

* Use int64_t type for number of elements.

* Support dense param and grad from OneDNN kernel.

* Enable SGD OneDNN kernel when use MP BF16 optimizer.

* Force non-copyable/movable OneDNNAXPYHandler.

* Reuse OneDNNAXPYHandler for spare tensors in SUM op.

* Fix SFINAE rules.

* Remove recording event inside AXPY.

* Get rid of internal primitive caching.

* Stop use PP cache mechanims to store mem and primitive obj.
* Handler obj store and reuse needed desc & prim

* Do not derive from MKLDNNHandlerT

* add layerwise learning rate for adamw

* fix format

* add unitest

* add NotImplementedError

* add gpu unitest

* update gpuplace