* add mish operator and api

* remove redundant code and modify grad_atol of mish unittest

* modify mish code to be consistent with other activation implementation

* modify comment of mish

* [Cherry-pick] Add the forward QR operator

* Add a check for multiplex op

* Improve multiplex based on reviews

* optimize backward (#37055)

* update

* update

* update

* modify code style

Co-authored-by: Nfeng_shuai <fengshuai03@baidu.com>

[CHERRY-PICK] Fix for matmul_v2 oneDNN op broadcasting when inputs dims have different lengths (#38733)

* fix for matmul_v2 broadcasting

* resolved conflicts

* support data_format='NHWC' for prelu channel mode (#37019)

* support data_format='NHWC' for prelu channel mode

* fix prelu weight shape for NHWC of static mode (#38310)

cherry-pick #36714

目前的fused_attention_op不支持attn_mask=None的输入，本PR对此进行了补充，并补充了相应的单测逻辑。

Add support for bias is none for fused_attention op.

* bug fix for  DeserializeSelectedRows. test=develop (#36520)

* fix SerializeSelectedRows (#36543)

* bug fix for  DeserializeSelectedRows. test=develop

* fix bug for SerializeSelectedRows. test=develop

* update. test=develop

* [Heterps]Refactor Heter Pipeline Parameter Server (#36845)

* change username

* fix

* fix

* fix

* fix

* fix

* update

* update

* update unittests

* fix

* update

* fix

* update

* fix

* fix

* fix

* update

* update. test=develop

* update. test=develop

* update. test=develop

* update. test=develop

* update. test=develop

* update. test=develop

* update. test=develop

* update. test=develop

* update. test=develop

* update. test=develop

* update. test=develop

* update. test=develop

* update. test=develop

* update. test=develop

* update. test=develop

* update. test=develop

* update. test=develop

* update. test=develop

* update. test=develop

* update. test=develop

* update. test=develop

* update. test=develop

* update. test=develop

* update. test=develop

* update. test=develop

* update. test=develop

* update send_and_recv op. test=develop

* update. test=develop

* update. test=develop

* update. test=develop

* update. test=develop

* update. test=develop

* update. test=develop

* update. test=develop

* update. test=develop

* fix. test=develop

* fix. test=develop

* fix. test=develop

* fix. test=develop

* fix ut. test=develop

* fix unit. notest,test=coverage

* fix ut. notest, test=coverage

* update. notest,test=coverage

* fix ut. notest, test=coverage

* fix ut. notest, test=coverage

* fix. notest, test=coverage

* fix. notest, test=coverage

* fix ut. notest, test=coverage

* fix ut. notest, test=coverage

* fix ut. notest, test=coverage

* fix ut. notest, test=coverage

* add func. notest, test=coverage

* fix ut. notest, test=coverage

* fix. test=develop

* fix. test=develop

* Fix unit test for send_and_recv_cpu & send_and_recv_gpu (#37129)

* [heterps]fix ut for heter_pipeline_trainer.cc  (#37136)

* fix ut. test=develop

* fix ut. test=develop

* [heterps]bug fix for local training with --heter_worker_num (#37166)

* fix. test=develop

* fix. test=develop

* fix. test=develop

* fix. test=develop

* fix. test=develop

* fix ut. test=develop

* fix ut. test=develop

* fix ut. test=develop

* [heterps]Refactor heterogenous worker (#37244)

* fix. test=develop

* fix. test=develop

* fix. test=develop

* fix. test=develop

* fix. test=develop

* fix ut. test=develop

* fix ut. test=develop

* fix ut. test=develop

* refactor heter trainer. test=develop

* fix. test=develop

* fix ut. test=develop

* fix ut. test=develop

* fix ut. test=develop

* fix ut. test=develop

* fix ut. test=develop

* fix ut. test=develop

* fix ut. test=develop

* fix ut. test=develop

* fix. test=develop

* fix. test=develop

* fix. test=develop

* fix. test=develop

* fix ut. test=develop

* fix ut. test=develop

* fix ut. test=develop

* [heterps]add heterps mode judgement (#37298)

* [heterps]change default executor for heter trainer (#37314)

* fix pslib. test=develop

* add device to train_from_dataset. test=develop

* refine fleet.stop_worker. test=develop

* fix ut. test=develop

* fix ut. test=develop

* fix executor & ut. test=develop

* fix executor & ut. test=develop

* fix executor & ut. test=develop

* [heterps]remove api for heter pipeline ps (#37396)

* fix api. test=develop

* fix api. test=develop

* fix code style. test=release/2.2

* fix CMakeLists. test=develop (#37454)

* enhance scatter err msg check

* fix ci error

* Add paddle.incubate.graph_send_recv API

* fix bug in CudaAtomicMin and CudaAtomicMax

* add empty line

fix bug to support dropout eval grad computing. cherry-pick #37305.

修复了fused_transformer_encoder_layer fine-tune过程发现的一些问题：

    fused_attention_op添加attn_mask=None的支持：PR
    pre_layer_norm处理问题：PR
    参数处理，计算错误的问题：PR
    add_bias计算错误问题：PR
    添加pure fp16的支持：PR

* add mlperf optimization PRs

* update

* fix rnn grad bug when num_layers is set 2 and dropout_prob is set 0

* add more test for rnn

    Renamed the variable and function
    Removed the original template function
    Removed the tests_properties in CMakeLists.txt

* fix cusparse compile bug in CUDA11.2, test=develop

* fix bug

Cherry-pick PR #36511

* Revert "Align CTC grad scale same with ESPNet (#34729)"

This reverts commit 10f9644c.

* ctc grad compute on gpu

* Fix bug when pre_layer_norm is false.

* update fft api path (#36219)

* update fft api path
* add sample code for ihfft2
Co-authored-by: Nchenfeiyu <chenfeiyu@baidu.com>

* fix fft axis (#36321)

fix: `-1` is used when fft's axis is `0`

* use unified external error message for cufft api (#36114)

* fft: modify sample code result (#36325)

* dynamic load mkl as a fft backend when it is avaialble and requested (#36414)

* add rocm support for fft api (#36415)

* move signal apis

* move fft and signal API path (#2)

* move signal apis

* move fft.py and signal.py to paddle/, fix typos

* fix relative imports from fft.py and signal.py

* fix typos in signal.py (#3)

* move signal apis

* move fft.py and signal.py to paddle/, fix typos

* fix relative imports from fft.py and signal.py

* fix typos

* disable Cache when CUFFT_VERSION >= 10200 (#4)

* move signal apis

* move fft.py and signal.py to paddle/, fix typos

* fix relative imports from fft.py and signal.py

* fix typos

* Add LRUCache for fft plans

* add LRUCache for cuff and hipfft (#5)

* move signal apis

* move fft.py and signal.py to paddle/, fix typos

* fix relative imports from fft.py and signal.py

* fix typos

* WIP: add cache

* delete move constructor and operator= for CuFFTHandle and FFTConfig

* remove log from CuFFTHandle and FFTConfig

* add lrucache for fft rocm backend

* disable LRUCache when CUFFT_VERSION >= 10200

* disbale copy and move for hipFFTHandle; format code
Co-authored-by: NXiaoxu Chen <chenxx_id@163.com>

* remove debug message of cufftHandler

* roll_op: support Tensor as input for shifts (#36727)

* fix fftshift/ifftshift on static mode

* update roll_op version

* add more test cases for fftshift/ifftshift
Co-authored-by: Nzhiboniu <31800336+zhiboniu@users.noreply.github.com>
Co-authored-by: Nchenfeiyu <chenfeiyu@baidu.com>
Co-authored-by: LJQ❤️ <33169170+lijiaqi0612@users.noreply.github.com>

功能：本PR的目标是提高attention模块的计算性能。
为了减少框架层对op的调度开销，本PR通过在C++层手动实现attention模块，对外提供attention 大op；
为了减少防存开销，本PR采取了两种优化方法：
（1）在q,k,v计算时通过共享输入X，将该处的gemm，transpose和bias add从三次调用减少为一次；
（2）使用kernel融合优化技术，在不同cuda kernel之间通过寄存器传输数据；

* Add FasterTokenizer Operator (#34491)

Add Tokenizer related functionalities for Transformer model in order that the process of training and predicting is consistent.

* support the text string as an input Tensor
* support the "VOCAB"unordered_map<wstring, int> as an input Tensor to lookup tokens
* Tokenizer used for BERT. This tokenizer applies an end-to-end, text string to wordpiece tokenization.
* It first applies basic tokenization, followed by wordpiece tokenization.

* optimize fast tokenizer

* remove const_cast
Co-authored-by: Nzhoushunjie <zhoushunjie@baidu.com>
Co-authored-by: Nwawltor <fangzeyang0904@hotmail.com>

* add op: fused_feedforward(backward) (#35611)

这个PR是fused_feedforward反向的代码

相关kernel实现：fused_dropout_act_bias, fused_residual_dropout_bias, fused_layernorm_residual_dropout_bias

fused_feedforward是一个融合算子，该算子对transformer模型的feed forward层的算子进行融合和封装，使得前端只呈现一个接口，通过融合减少部分访存和kernel launch的时间，以此提升性能。

* Move fused_attention and fused_feedforward functional api path to incubate (#36704)

将 #35905 和 #35843 PR中新增的的python api接口移到incubate目录下。

This is a fusion operator to compute feed forward layer in transformer model architecture.