* refine layer to

* delete comment

* refine logic

* refine code

* refine pure_fp16_init

* refine comment

* try to fix CUDA Graph H2D copy bug

* remove useless code

* fix ci

* fix ROCM CI

* fix CUDA_VERSION

* improve CI coverage

* graph engine demo

* upload unsaved changes

* fix dependency error

* fix shard_num problem

* py client

* remove lock and graph-type

* add load direct graph

* add load direct graph

* add load direct graph

* batch random_sample

* batch_sample_k

* fix num_nodes size

* batch brpc

* batch brpc

* add test

* add test

* add load_nodes; change add_node function

* change sample return type to pair

* resolve conflict

* resolved conflict

* resolved conflict

* separate server and client

* merge pair type

* fix

* resolved conflict

* fixed segment fault; high-level VLOG for load edges and load nodes

* random_sample return 0

* rm useless loop

* test:load edge

* fix ret -1

* test: rm sample

* rm sample

* random_sample return future

* random_sample return int

* test fake node

* fixed here

* memory leak

* remove test code

* fix return problem

* add common_graph_table

* random sample node &test & change data-structure from linkedList to vector

* add common_graph_table

* sample with srand

* add node_types

* optimize nodes sample

* recover test

* random sample

* destruct weighted sampler

* GraphEdgeBlob

* WeightedGraphEdgeBlob to GraphEdgeBlob

* WeightedGraphEdgeBlob to GraphEdgeBlob

* pybind sample nodes api

* pull nodes with step

* fixed pull_graph_list bug; add test for pull_graph_list by step

* add graph table;name

* add graph table;name

* add pybind

* add pybind

* add FeatureNode

* add FeatureNode

* add FeatureNode Serialize

* add FeatureNode Serialize

* get_feat_node

* avoid local rpc

* fix get_node_feat

* fix get_node_feat

* remove log

* get_node_feat return  py:bytes

* merge develop with graph_engine

* fix threadpool.h head

* fix

* fix typo

* resolve conflict

* fix conflict

* recover lost content

* fix pybind of FeatureNode

* recover cmake

* recover tools

* resolve conflict

* resolve linking problem

* code style

* change test_server port

* fix code problems

* remove shard_num config

* remove redundent threads

* optimize start server

* remove logs

* fix code problems by reviewers' suggestions

* move graph files into a folder

* code style change

* remove graph operations from base table

* optimize get_feat function of graph engine

* fix long long count problem

* remove redandunt graph files

* remove unused shell

* recover dropout_op_pass.h

* fix potential stack overflow when request number is too large & node add & node clear & node remove

* when sample k is larger than neigbor num, return directly

* using random seed generator of paddle to speed up

* fix bug of random sample k

* fix code style

* fix code style

* add remove graph to fleet_py.cc

* fix blocking_queue problem

* fix style

* fix

* recover capacity check

* add remove graph node; add set_feature

* add remove graph node; add set_feature

* add remove graph node; add set_feature

* add remove graph node; add set_feature

* fix distributed op combining problems

* optimize

* remove logs

* fix MultiSlotDataGenerator error

* cache for graph engine

* fix type compare error

* more test&fix thread terminating problem

* remove header

* change time interval of shrink

* use cache when sample nodes

* remove unused function

* change unique_ptr to shared_ptr

* simplify cache template

* cache api on client

* fix
Co-authored-by: NHuang Zhengjie <270018958@qq.com>
Co-authored-by: NWeiyue Su <weiyue.su@gmail.com>
Co-authored-by: Nsuweiyue <suweiyue@baidu.com>
Co-authored-by: Nluobin06 <luobin06@baidu.com>
Co-authored-by: Nliweibin02 <liweibin02@baidu.com>
Co-authored-by: Ntangwei12 <tangwei12@baidu.com>

* Refactor conv2d int8 unit test

* Correct according to review and add int8 check

* refine vlog of interpretercore

* fix ut

* initial tensor design & sign kernel demo

* add move constructor for meta & add lodtensor

* add dirs & sign xpu kernel

* add mean cpu&cuda kernel impl

* move sign & mean xpu & npu kernel

* add selected_rows basic impl

* refactor design, BaseTensor to DenseTensor, etc.

* add scale mkldnn kernel

* polish xpu & npu impl details

* fix mkldnn reuse compile failed

* change tensor operation lib name

* rename util filename

* add more comments

* change TensorImplInterface to TensorInterface

* add kernel key and factory

* remove MKLDNNTensorMeta, add MKLDNNDenseTensor

* change XXDeviceContext to XXContext

* add base kernel registrar utils & test on sign

* replace boost::any by paddle::any

* fix several ci failed

* fix npu compile error

* add ordered map util

* fix multiple ordered_map compile errors

* move dev into include dir

* support sign op in static op run

* fix static op run error

* fix new executor compile failed

* add dygraph branch & remove sign_op.h

* fix test_infer_no_need_buffer_slots

* fix rocm compile link error

* fix unitybuild error & clear glog

* fix npu compile failed

* skip quant trans test

* fix part windows compile problem

* fix xpu enforce error

* fix inference test failed

* remove ordered_map to solve quant failed

* fix part of rcom compile faild

* add more register kernels

* revert scale kernel temporarily

* fix code format error

* add new kernel registrar marco

* rename top to tcmpt

* revert xpu, npu, mkldnn impl & remove op def

* add kernel args parse functor to auto parse args

* revert some change & add scale kernels

* add op proto in dygraph kernelcontext building

* polish kernel dispatch logic & nameing rule

* fix scale kernel match error

* fix scale test failed

* add mean API and unittest

* test mean api success

* add branch to solve compiled error

* skip clang format error

* add mean skip rule in op_library

* add dot kernel, api and unittest (#6)

* remove old kernel and add symbol link

* fix dot compiled failed

* add merco for module declare

* fix npu and xpu compile error

* revert sign, mean, scale, dot kernel removing

* add comment for keeping old kernel impl

* fix mutable_data error

* fix bfloat16 conflit

* fix inference undef error

* adapt to msvc compile rules

* polish comment for template inst

* add cmake template instantiation for win

* fix backend to place device id bug

* fix ifdef error

* Op2functor (#7)

* add kernel args maker class

* make args maker non-const

* remove debug log

* modify codes by review options

* split constructPrKernelContext function

* fix output name bug

* fix test_mean_op test_sign_op failed

* fill_any_like kernel refactor (#10)

* fill_any_like kernel refactor

* remove useless code of full_like c++ api

* skip dtype for fill_any_like

* add attrs for kernel key constrcut

* add use_pt_kernel Flags to control whether to use pt kernel (#13)

* add use_pt_kernel Flags to control whether to use pt kernel

* change the default value to true for cheking pt kernels

* fix mutable_data cuda place error

* move high level apis into hapi

* remove selectedrows adapting temporarily

* Support Scalar in Tensor Compute Library (#14)

* fill_any_like kernel refactor

* remove useless code of full_like c++ api

* Support Scalar in Tensor Compute Library

* add scalar in dygraph and static graph mode

* keep the basic type for attr, instead of using scalar for all

* merge the code

* remove mkldnn tensor & polish details

* use flat_hash_map and small_vector in kernel factory

* Refactor flatten kernel (#12)

* refactor flatten kernel

* update infershape function

* fix compile bugs

* fix bugs when merge

* fix compiler bugs

* fix bugs when run test_flatten_api

* fix bugs when run test

* Revert "use flat_hash_map and small_vector in kernel factory"

This reverts commit 23091495cfdd3df8cc1be592d30f09ea66a7c72b.

* Move cpu, cuda and other device code into kernels (#15)

* fill_any_like kernel refactor

* remove useless code of full_like c++ api

* Support Scalar in Tensor Compute Library

* add scalar in dygraph and static graph mode

* keep the basic type for attr, instead of using scalar for all

* merge the code

* start refactor matmul

* move cpu, cuda and other device modules into kernels

* merge code

* polish code in operator.cc

* Perfect unitests (#16)

* perfect unittest

* update license

* replace with flat_hash_map, small_vector (#19)

* fix small_vector build error on windows platform

* replace with flat_hash_map, small_vector

* remove todo

* Perfect unitests (#20)

* perfect unittest

* update license

* fix bug when run tcmpt_utils_test

* refactor execution adapting impl

* fix insert conflit

* Fix CI bug of test_yolov3 (#21)

* fill_any_like kernel refactor

* remove useless code of full_like c++ api

* Support Scalar in Tensor Compute Library

* add scalar in dygraph and static graph mode

* keep the basic type for attr, instead of using scalar for all

* merge the code

* start refactor matmul

* move cpu, cuda and other device modules into kernels

* merge code

* polish code in operator.cc

* Fix CI bug of test_yolov3

* add the tensor base class, test=develop (#17)

* update the tensor base class, test=develop

* remove two funcs, test=develop

* update the error msg, test=develop
Co-authored-by: NChen Weihang <chenweihang@baidu.com>

* [no-verify] commit backend and tensor signature changes

* Rename tcmpt to pten (#23)

* rename tcmpt to pten

* update omitted files for rename to pten

* update omitted file for rename to pten

* remove k of all enum var

* remove kernel_instantiate (#26)

* remove symbols and spatial_tensor

* change common to functions

* readd share tensor impl methods

* add a candidate dense tensor class, test=develop (#28)

* change all Pt to Pten

* resolve conflit with xiaowei

* Op2functor opt1 (#27)

* replace to small vector and change to const &

* add std::move
Co-authored-by: NChen Weihang <chenweihang@baidu.com>

* polish kernel factory and kernel registry

* fix operator test error msg mismatch

* remove tensor signature and backend set member

* move scalar and polish enforce

* revert dtype layout change to fix error

* fix enum operator override error

* add several base unittests

* add pten utils tests

* polish some details

* Dev/op2func refactor 3 (#30)

* add a candidate dense tensor class, test=develop

* remove TensorBase::backend(), test=develop

* remove some ops, test=develop

* cherry-pick the pr of tensor meta, test=develop

* moves the dense tensor and some ops, test=develop

* update the linalg operator, test=develop

* update other operators, test=develop

* fix errors, test=develop

* fix bugs, test=develop

* try to resolve the problem of windows ci, test=develop

* updates codes, test=develop

* fix the tensor_utils.cc, test=develop

* modify the dense tensor, test=develop

* fix the data type, test=develop
Co-authored-by: Nshixiaowei02 <39303645+Shixiaowei02@users.noreply.github.com>

* polish some details

* polish kernel signature details

* fix a bug about offsets of the tensor, test=develop (#31)
Co-authored-by: Nshixiaowei02 <39303645+Shixiaowei02@users.noreply.github.com>

* polish some details
Co-authored-by: Nchentianyu03 <ctychentianyu@gmail.com>
Co-authored-by: Nzyfncg <1370305206@qq.com>
Co-authored-by: NYuanRisheng <yuanrisheng@baidu.com>
Co-authored-by: N石晓伟 <39303645+Shixiaowei02@users.noreply.github.com>

* fix some bug in new executor, test=develop

* fix error message, test=develop

* add trainer desc config to distributed strategy

* code style modified

* data_feed set lod

* fix bug

* code style

* fix bug

* save load

* save load

* save unittest

* add unittest of the_one_ps

* unittest

* add todo in communicator sendsparse

* Add bincount op

* upload cpu version

* fix unitest

* fix unittest

* fix unittest

* fix en doc

* add more test

* fix en doc

* add more test case

* fix test

* fix input vailidation

* fix input check

* fix unittest

* fix test

* fix en doc

这个PR只包含fused_feedforward前向的代码。

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。

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；

* [hapi] support dygrapg amp O2

* fix problem of static pure fp16 in hapi

* fix bug

* fix format

* fix ut

* follow comments

* update ut

* update amp save/load

* fix ut

* refine code format

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.

* update

* fix ut error

* update ut

* fix async_read bug

* change index place to cpu

* add tensor size judge

* add async_read & async_write test

* fix bug in async_write

* fix mac py3 ci

* fix bug for cpu version paddle

* fix windows ci bug

* change input argument error type

* change const_cast to mutable_data

* add async_write out-of-bound check and consumate error hint

* fix a small bug for dst_tensor

* add docs and refine codes

* refine docs

* notest,test=windows_ci

* fix windows ci

* fix require

* fix code-block

* add core.is_compiled_with_cuda()

* refine amp level

* fix typo

* update tracer._amp_level

Remove RunFromCinn method in PE because We Will Call CinnRunner in Compute method of SubgraphOp

Add use_cinn flag and use it to control whether we run PaddlePaddle using CINN.

Also add:

Replace PaddlePaddle graph with a CINN graph in a pass
PE Method to feed data and run the graph by CINN

* support CUDA Graph on PE

* add ut, fix CI compile

* reduce memory consumption

* fix CUDA 10 CI

* improve coverage

* improve python coverage

Add python interface of subgraph: 1. all_sub_graphs() 2. get_sub_graph(idx)

* add basic support for CUDA Graph

* fix ci compile error

* fix LOG print, fix windows CI

* follow comments and update

* small fix for default ctor

* fix rocm compile error

* fix CPU compile error

* Initial Commit

* add unittest and add error information

* modify doc

* fix some error

* fix some word

* fix bug cudaDeviceProp* and modify error explanation

* fix cudaDeviceProp* error and unnitest samples

* fix hip error and PADDLE_WITH_HIP

* update style

* fix error is_compiled_with_cuda

* fix paddle.device.cuda.get_device_properties

* fix error for multi thread safe

* update style

* merge conflict

* modify after mentor review

* update style

* delete word

* fix unittest error for windows

* support string input and modify some code

* modify doc to support string input

* fix error for express information

* fix error for express information

* fix unnitest for windows

* fix device.startswith('gpu:')

* format error and doc

* fix after review

* format code

* fix error for doc compile

* fix error for doc compile

* fix error for doc compile

* fix error for doc compile

* fix error for doc compile

* fix py2 error

* fix wrong words and doc

* fix _gpuDeviceProperties

* add func/class API psroi_pool and UT

* add UT in static mode

* Remove redundant type checks in static mode

* More detailed description for test_psroi_pool_op

* fix code format of UT

* fix en-doc

* set file_num in one shard

* format

* adam to adamw in AdamW

* add lr_ratio in adamw

* refine logic bug in cpu adamw

* delete fix bug for cpu adamw

* delete fix bug for cpu adamw

* Fix copy elision warning

* Remove redundand code

Add basic Cost Model, it uses executor to run program and profile it to get op time.

This is an early basic version, we will add more functions in the future.

* split cuda_profiler into .h and .cc

* fix cmake

* remove inline

* Clean ParaseMemInfo and fix unittest with multi-thread

* fix declare

* add pure fp16 major function in auto_cast & tracer

* support master weight in dygraph for pure fp16

* check mix dtype of fp16&fp32 for check_finite_and_unscale op

* change pure fp16 funtion name

* refine some bug in auto_cast

* refine auto_cast interface logic

* add param _casted_by_pure_fp16 for class Layer

* support state_dict hook for save model by user appointed dtype in pure_fp16_decorator

* refine pure_fp16_decorator as decorator

* add unittest

* add comment

* add comment

* support recompute

* add comment for auto_cast and decorator

* support to_static_state_dict for paddle.jit.save

* unlimite models num and optimizers num

* add lookup_table in black_list

* fix momentum and layer state_dict

* fix bug in layer state_dict

* fix bug in layer state_dict_helper

* refine unittest

* refine test_momentun_op

* refine interface and some code

* refine amp_decorator interface

* refine pure fp16 interface

* refine master weight interface