* extract common methods to reuse

* add header for transpose ops

* fused_transpose

* Split big function

* transpose2 tests

* fused_transpose

* Apply extra attributes

* add pbtxt file

* update pbtxt

* Merge develop

* add more strict op compats

* code  style

* remove mkldnn_data_type

* unify SetOutMemDescWithReshape2FuseSupport

* adjust quantize-dequantize for transpose

* remove appendact

* transpose2 quantization

* fix int8 tests

* adjust transpose_op to current develop

* delete fusion code from transpose_kernel

* add fused transpose to NHWC unittest

* change order

* move DataType from paddle::experimental to phi

* convert namespace

* convert namespace

* convert namespace

* clarify namespace

* convert more datatype

* Revert "convert more datatype"

This reverts commit 083b462959e6a22d4d8767707b628b95b396642e.

* convert more in auto_code_generator

* fix conflicts for XPU

* fix namespace conflicts

* fix errors

* Revert "fix errors"

This reverts commit f9d9958b54ee32141112274c8a5c3c381ab0f876.

* fix errors

* fix formatting

* split pass

* fix compile

* fix ut

* more time

* modify ut

* reduce dim

* fix compile

* reshape weight

* tensor

* remove enforce

* static shape ut

* batchsize

* reorder pass

* minus test cases

* windows timeout

* windows time out

* remove test for windows

* correct

* sssss

* xxx

* Revert "Revert "【Hackathon No.67】remove operator.h in blas.h (#50989)" (#51467)"

This reverts commit b9d91531.

* remove cout

* add header

* fix missing header

* fix refer fluid error

* fix missing header

* 更新 repeat_interleave_grad_kernel_impl.h

Change to phi style datatype.

* 更新 repeat_interleave_grad_kernel_impl.h

Fix missing header

* datatype fluid -> phi

* paddle::experimental -> phi

* fix reference error

* fix reference error

* fix reference error

* fix errors

* fix missing FLAGS

* fix missing headers

* fix missing headers

* fix missing headers

* fix missing headers

* fix missing header

* fix missing header

* fix errors

* mkldnn->onednn

* fused softplus op + kernel

* remove extra attributes

* add missing handler

* change var name

* use python to generate cutlass code

* refine CommonConvKernelPart1, CommonConvKernelPart2

* remove useless code in generate_cutlass_code.sh

* add more config in conv2d_residual

* CommonCutlassConvKernelPart1 and CommonCutlassConvKernelPart2

* add group conv support in util.cu

* remove .sh

* refine name

* make name goodgit status!

* add fuse_alpha

* make code easy to understand

* mot fopen generate in py

* use python script to generate conv2d,group=1 cutlass code

* use const &

* use const & && use python script to generate conv2d/group=1 code

* first approach

* test finished

* cpp test deleted

* CmakeList corrected

* multi_gru_seq_fuse_pass rewritten

* dummy cout deleted

* review changes

* timeout extended

* fix performance drop in BF16 models

* fix test_cpu_quantize_squash_pass

* Added file

* Tests separated and rewritten, fixed fc_lstm_fuse_pass

* Resolve conflicts

* forbid tensorrt_engine op's output is a persistable var

* add sm version check

* use GetGPUComputeCapability

* ConvertToFusedOp

* change static to inline
Co-authored-by: NTomasz Socha <tomasz.socha@intel.com>

---------
Co-authored-by: NTomasz Socha <tomasz.socha@intel.com>

* support bw invoke fw

* fix scale in static_backward.yaml

* fix the bug in tensorrt/convert

* move 'scale','sign' into ops.yaml

* add scale_grad of scale in op_compat.yaml

* change generated_static_op in CMakeLists.txt

* change SHA

* update to oneDNN 2.7

* update to 2.7.1

* update to 2.7.2

* add supported hardsigmoid

* update to 2.7.3

* limit cpu threads for int8 test

* group activations

[XPU] add multi_encoder_xpu_slice_fuse_pass, generate_sequence_xpu_fuse_pass, generate_sequence_xpu kernel (#50570)

* Add matmul_v2 to the quantization process and adjust Ernie model test

* Correct cpu_quantize_pass test

* Move op to fuse transformation to placement pass

* Correct test

* New onednn test

* checkopoint

* added new test, fixed issue with onednn bias

* fix bias check

* remove prints, refactor code

* delete old test

* update python tests cmake

* Delete depracated conv bias

* Delete outdated bias from convolution test

* support xpu multi-card infer

* add ut

* clean code

* clean code

* fix

* fix

* fix

* fix

* Done

* Deleted old python test, fixed new python test, changed names in parallel_UT

* Revert parallel UT changes

* Revert parallel UT changes v2

* Review fixes and simplification of conv output shape calculation, disabled sqrt from conv_act_duse_pass

* delete sqrt from possible activations from conv_concat_relu test

* review refactor

* merge main

* delete sqrt from list of compatible activations

* Test with no outdated inputs

* move profiler

* add file

* fix mac compile bugs

* fix ci bugs

* fix mac bugs

* fix ci bugs

* fix compile bugs

* perfect code according comment

* expand mix_precision to custom_device

* fix bug

* fix bug

* fix comment

* fix DEFINE bug

* upgrade protobuf to 3.19.0 in cmake

* recover protobuf python version

* fix distribute compile

* fix

* fix framework.data_feed_pb2

* fix macos ifdef

* fix lite

* test

* update protoc from 3.19.0 t0 3.20.0

* test

* debug

* test

* test

* debug

* debug

* debug

* debug

* test

* debug

* update protocol from 3.20.0 to 4.21.12

* modify graph_brpc_client.h

* modify graph_brpc_client.h

* test

* test

* test

* fix third_party cache problem on build ci

* updata proto

* test

* test

* test

* test

* test

* test

* fix coverage failed test

* try to fix test_exe_fleet_model_run

* fix cinn bug

* fix windows compile problem

* fix python/requirements

---------
Co-authored-by: Npangyoki <pangyoki@126.com>