* fix expand as op

* fix bug

* group norm fp16 support

* add bfloat16 support for more ops

* fix ci compile

* fix windows compile error

* fix windows compile error

* fix rocm compile error

* fix ROCM compile error

* add index sample fp16 support

* remove fluid APIs in distributed_strategy.py and role_maker.py

* Revert "remove fluid APIs in distributed_strategy.py and role_maker.py"

This reverts commit 223bbee990d3bf69e252fc3c0f19e3873550a264.

* remove fluid APIs in distributed_strategy.py and role_maker.py

* remove index sample op changes

* remove fluid APIs under fleet.base

* remove fluid APIs under fleet.layers.mpu

* remove fluid APIs under fleet.meta_optimizers

* fix fluid error

* fix util_factory.py

* reset fluid.io.load_inference_model API

* rm dependence to "convert_utils.h" in some files

* fix bugs

* replace DataType2String with DataTypeToString

* replace framework::DataTypeSize with phi::SizeOf

* mv convert_function from fluid to phi and rm old map

* recommit with pre-commit

* repalce ProtoVarType with ProtoDataType and update comment.

* fix error about include "dnnl.hpp"

* revert add dep mkldnn to convert_utils in phi

* add mkldnn deps in convert_utils.h in phi

* move deps to convert_utils.h in phi

* do not calc reduce_all in eager mode

* refine python c cast list

* refine

* refine

* refine

* refine

* refine

* refine

* refine

* refine

* refine

* decouple im2col from fluid

* move im2col to phi

* fix build error

* delete redundant comment

* add stat tool

* add roll and roll_grad kernels and strided_slice and strided_slice_grad kernels, test=kunlun

* add masked_selected_grad kernel,test=kunlun

* first commit

* 2nd commit

* Migrate elementwise_div

* Migrate elementwise grad kernels

* move vol2col from fluid to phi

* update copyright year

* cleanup unused code

* unify is_int8 is_bfloat16

* Simplify matmul_v2 FWD kernel

* remove RunKernel methods

* remove import namespace

* remove headers

* clean fluid/phi cross imports

* remove fluid axpy_handler

* delete fluid methods

* activations

* OneDNNMemDesc

* MKLDNNFormatForSize

* MatchShapeToLayout

* MKLDNNMemoryFormat

* MKLDNNFormat

* ReorderMKLDNNHandler

* to_void_cast

* review suggestions

* interpolate

* remove fluid depedency

* init

* ExecuteMatMulV2

* rm fluid kernel

* matmul_grad

* remove mutable_data

* mul_grad

* refine reduce_all

* Fix wrong eigen header include

* fix compile bug

* move threadpool

fix cmake

* fix make

* move cross_entropy from fluid to phi

* replace mutable_data with Alloc

* use .template

* cleanup unused code

* unify is_int8 is_bfloat16

* Simplify matmul_v2 FWD kernel

* remove RunKernel methods

* remove import namespace

* remove headers

* clean fluid/phi cross imports

* remove fluid axpy_handler

* delete fluid methods

* activations

* OneDNNMemDesc

* MKLDNNFormatForSize

* MatchShapeToLayout

* MKLDNNMemoryFormat

* MKLDNNFormat

* ReorderMKLDNNHandler

* to_void_cast

* review suggestions

* interpolate

* remove fluid depedency

* init

* ExecuteMatMulV2

* rm fluid kernel

* matmul_grad

* remove mutable_data

* Migrate conv_transpose to phi

* Move handler to kernel

* kernel m

* Fix formatting

* handler

* remove fluid

* revert tcp_store

* tcp_store

* remove unused

* Fix declaration

* add dnn input

* Fix typo
Co-authored-by: NSławomir Siwek <slawomir.siwek@intel.com>

* Add quick gelu and fused bias add kernel

* fix annotation

* remove useless code

* add fast gelu option and set it in multi transformer op

* add flag to restrict if use fast gelu approximate

* fix flags conflict

* fix use tanh function instead

* add cudart version limit

* use phi fast tanh func

* fix comment

* move "paddle/phi/backends/gpu/gpu_device_function.h" to phi

* update copyright years

* rm "fluid/platform/device/gpu/gpu_device_function.h" in phi

* fix rocm-complie bugs

* correct sync behavior for XPU distributed training

XPU support event mechanism similar to cuda event, so it is advisable to
use an event to sync compute/comm streams for performance. However this
mechanism is never fully tested, and inconsistent loss/ending_epochs are
reported. Therefore, this PR replaces event sync with stream waiting as
a temporary solution.

* remove compile warning

* Refactor conv_kernel and conv_grad_kernel to provide interface for CUDNNv8 implementation

* Fix macro

* Add implementation for conv_kernel and conv_grad_kernel

* Modification after rebase onto latest develop

* Modify plan cache to comply with the API of phi::autotune

* Refactor to reduce duplicate code

* Review fix:
- move functions in  conv_kernel_impl_v8.h and conv_grad_kernel_impl_v8.h to conv_kernel.cu and conv_grad_kernelk.cu
- add const specifier for input tensor
- add logging when plans fail to execute
- move CudnnConvBwdFilterV8 and CudnnConvBwdDataV8 to conv_cudnn_frontend.h

* - move plan building outside of cache

* Fix ROCM build

* [NPU] add _npu_identity op and api, test=develop

* fix doc

* address comments