Add paddle.geometric.weighted_sample_neighbors API

* remove kSPARSE_WEIGHTS

* remove kFASTER_DYNAMIC_SHAPES_0805 and add 'TrtMajorVersion' function

* [Zero-Dim] update 0d tensor api en doc, test=document_fix

* [BUG] fix windows kernel dispatch of _lzcnt bug (#53728)

[AMP] Allow to switch whether to use promote strategy to choose kernel for O2 training. (#53742) (#53841)

Pcard-70458

cherry-pick #53742

中文文档：PaddlePaddle/docs#5882

* delete log

* filter some kernel signature

Pcard-70458
cherry-pick: #53770

Pcard-70458
cherry-pick #53712

* scale, square, sum, swish trt op converter support zero dim (#53660)

* [Paddle-Inference] Support trt 0dims of expand_as_v2 and mish. (#53627)

* support_expand_mish

* add unitest for reshpe 0 dims (#53685)

* Add trt pow converter. (#53462)

* Add trt pow converter.

* update to use AddConstantLayer

* add dims=0 ut

* [inference Zero-Dim]add equal, elementwise_op trt 0d (#53704)

* [inference Zero-Dim]prelu trt converter support zero dim tensor (#53634)

* prelu op trt converter support zero dim

* [Inference Zero-Dim] Support trt 0dim of gelu, hard_swish, hard_sigmoid and leaky_relu (#53714)

* support_act
* delete_silu

* [inference zero dim] softmax, stack op trt converter support zero dim (#53729)

* softmax support

* support stack

* remove unused code

* update

---------
Co-authored-by: NYuanle Liu <yuanlehome@163.com>
Co-authored-by: Nxiaoxiaohehe001 <49090790+xiaoxiaohehe001@users.noreply.github.com>
Co-authored-by: Nzhoutianzi666 <39978853+zhoutianzi666@users.noreply.github.com>
Co-authored-by: NWilber <jiweibo@baidu.com>

This PR fix docs error of index_put , related dev PR is #53727

This PR add data_type for selecting which arg's datatype to instantiate template type T for index_put kernel
Related PR #53652

* Fix div error when dtype is int64 in static mode

* Fix out dtype

Fix static_assert bug in Windows CUDA 11.6 compilation. This may be the bug of msvc.

【cherry-pick】【BugFix】fix err of api `to_tensor`, which caused by numpy version update (#53534) (#53624)

* 【BugFix】fix err of api `to_tensor`, which caused by numpy version update (#53534)

* fix

* update code

* pre-commit

* remove scale check (0-D tensor is usable)

* fix data dtype err

* fix numpy default dtype diff

* fix data dtype

* fix data dtype

* update

* fix coverage

* fix old test which is not correct when 0-D tensor is usable

* up warning level

* numpy still vlog-0

This PR add index_put api for paddle

cherry-pick #53659

Fix bug in log_softmax kernel when lastdim is larger than 100000

There is an unexpected log in the calculation

Cherry-Pick: #53654

Revert argsort to the version without full sort algorithm implemented

* Support different dtypes of inputs for broadcast for dropout optimization  (#52093)

* change judgement for DropoutGradGPUKernelDriver

* add UnrollerWithoutVecSize and after this Loaddata to be refined

* pass unittest

* use same unroller with XPU

* BroadcastWithInt64Index

* BroadcastDataLoader template partial specialization

* fix compile errs in ROCms

* PR comment

* dropout_nd_optimization (#51479)

* with printf

* add DropOutNdForwardKernel

* PR comment

* Dropout optimize & clean broadcast inT and ElementwiseType (#52969)

* change judgement for DropoutGradGPUKernelDriver

* add UnrollerWithoutVecSize and after this Loaddata to be refined

* pass unittest

* use same unroller with XPU

* BroadcastWithInt64Index

* BroadcastDataLoader template partial specialization

* fix compile errs in ROCms

* clean ElementwiseT and InT for BroadcastKernel

* default axis and clean inT

* remove redundant fast divmod computation

* optimize drop_nd & drop_nd_grad

* optimize BroadcastDataLoader bf16 fp16

* rm InT etc. after merge develop

* delete constexpr for windows ci

* fix conflict

* fix conflic with develop

* fix conflic

* new clean

* clean

* Fix xpu2 kp compile error (#53548)

* fix conflict

* conflict

[Cherry-pick 2.5][Zero-Dim]  paddle.static.data, squeeze, unbind, unstack, gather_nd and einsum support 0D (#53602)

* add test cases, test=allcase

* fix test cases, test=allcase

* fix test cases, test=allcase

* assert_allclose, test=allcase

* 1e-5 to 1e-4, test=allcase

* change rtol from 1e-4 to 1e-3, test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* fix test cases, test=allcase

* fix test cases, test=allcase

* modify the test_squeeze to not use Tensor type axis, test=allcase

* add grad check for unbind and unstack, test=allcase

* check for squeeze axis tensor type, test=allcase

* fix bug, test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

fix static promote
将因性能有问题而放入unsupprot_list中的算子放入黑名单中，以保证在O2模式下，只有3种场景权重会保持fp32

cherry-pick #53430

* fix doc erros, test=allcase

* conflict

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* test=allcase

* fix doc erros, test=allcase

* fix the to_tensor error

Cherry pick fused linear

* [inference][trt]trt support 0 dims (#53383)

* trt support 0 dim
* update activation ut
* fix trt Unary operation do not support 0d when TRT < 8.6
* Update op_teller.cc
* update unary ut
* add rsqrt to unary_list
* move rsqrt to act_list

* add complex support for  optest

* add complex grad test

* append one

* move some debug info

* move some debug info

* move some debug info

* move some debug info

* add more complex test

* Fix naming ambiguity

* Revert "add more complex test"

This reverts commit dbcb0516b8e53ba42e2d6089878a39b395345969.

* change backward gradient, add TODO

* [Zero-Dim] fix functool.reduce more safe with intial value, to support empty list (#53182)

* [Zero-Dim] support 0d tensor for shape and squeeze onednn kernel (#52832)

* support 0d tensor for shape and squeeze onednn kernel

* set python api for shape op ut

* [Zero-Dim] distributed scatter/all_to_all support input 0D tensor (#53186)

* [Zero-Dim] Support paddle.sum/mean/loss api output 0D,test=allcase (#52739)

* [CINN Support 0D-Tensor] CINN supports 0D-Tensor with trick temporarily (#53382)

* [CINN Support 0D-Tensor] CINN supports 0D-Tensor with trick temporarily

* Add unittest

* [CINN Support 0D-Tensor] CINN hack squeeze2 with trick temporarily (#53454)

* fix test_autograd_dynamic (#53473)
Co-authored-by: Nzhwesky2010 <zhouwei25@baidu.com>

---------
Co-authored-by: NYangQun <qun.yang@intel.com>
Co-authored-by: NHongyuJia <jiahongyu@baidu.com>
Co-authored-by: NHydrogenSulfate <490868991@qq.com>

* support 0-D output and 0-D as indice in __getitem__

* fix tests

* fix inference and UT

* add unittest for setitem

* fix xpu test

* fix xpu 0-d

* fix right value is 0d and index is List/Tensor

* Hack__getitem__ from 0-d to 1-d with FLAGS_set_to_1d

* change PHI_DECLARE_xxx to DECLARE_xxx since the change not merged to 2.5

* hack 1-D tensor to Scalar

* throw warning at __getitem__, not slice_utils

* add ```converter_type``` for op converter

Cherry-Pick: #53582
修改内容：在除法out = x / y中，将y的反向公式由dy = -dout * out / y 改为 dy = -dout * ((x / y) / y)
修改原因：使用result作为反向的输入，在低精度的时候本身cast之后就会存在一些精度损失，所以重新计算后才是更准确的结果
修改影响：此改动可以使结果更精确且对性能影响忽略不计

修复优化器精度检查bug

* Add fused_gate_attention API. (#53432)
* Add PADDLE_THROW in take_along_axis kernel when the datatype of index is wrong. (#53556)