[Inference] Replace groupNorm when data types are bf16 and fp16, and data format is NHWC implementation. (#55399)

* finish

* cpergroup odd

* fix bf16

* single channel

* code style

* jingdu duiqi

* add head_file

* add bf16 head file

* bf16 2

* bf16

* bf16 head

* bf16 compile

* py test

* bf16 compile

* bf16 compile

* unset py test

* nhwc

* test

* mean var

* bf16 success

* su

* ctest success

* use is_same_as

* is_same

* use is_same

* rtol

* gpu_stream

* del sigmod

* fix bfloat16 type

* use cuda_bf16_hpp

* use_cuda_arch

* bfloat162float2

* del inplace_tol

* del max_releative_tol

* temp store

* jingdu duiqi

* temp store

* plugin

* jingdu duiqi

* duiqi

* include cuda.h

* del half

* half single

* ci

* add const

* ci

* cudamemset

* del printf

* fp16 test

* add half compute

* del br16 ci

* del ci

* ci approve

* del fluid include

* [Doctest]fix No.21, test=docs_preview

* Revert "[Doctest]fix No.21, test=docs_preview"

This reverts commit 76bcdb280e254d682be6fc6f85588f1940bb1ade.

* [Doctest]fix No.21, test=docs_preview

* fix bugs,test=docs_preview

* [Doctest]fix No.22-24,26,27, test=docs_preview

* update fix

* with pre-commit, test=docs_preview

* fix seed, test=docs_preview

* fix error, test=docs_preview

* fix seed, test=docs_preview

* fix seed, test=docs_preview

* simple reaplce

* for debug

* fix bugs

* fix some bugs

* del fill_constant_batch_size_like

support stride

【inplace api】batch add inplace api paddle.log_, paddle.i0_, paddle.nn.functional.leaky_relu_... (#55576)

* batch add inplace api

* add inplace test

* add activation inplace

* fix test

* remove atan2 ge, gt, le, lt, nq

* remove atan2 ge, gt, le, lt, nq

* fix windows ci error

* rerun ci

* fix typro

* fix bugs

---------
Co-authored-by: Nzhangrui34 <v_zhangrui34@baidu.com>

* remove api staticrnn

* move select_input/output to static/controw flow

* delete some func, only remain Switch

* clean fluid.layers.controw_flow

* remove fluid.layers.controlflow

* fix conditional_block ut

* [Kunlun] Mofify some legacy code on distributed training

There were limitations on XPUs before, such as concat/split is not
supported, and c_broadcast only support fp32. These limitations are
lifted recently.

Multi-device profiling on XPU will also be supported by this PR.
Without this PR, a hanging broadcast will be issued by devices that
enables profiling, eventually lead to kernel timeout error.

* fix typo

* [Add]Add Xdoctester

* [Fix]fix beta docstring

* [Doctest]change dirichlet docstring

* [Doctest]change gumbel docstring

* [Doctest]change bernoulli docstring

* [Doctest]change categorical docstring

* [Doctest]change ops.py docstring

* [Doctest]change conv docstring

* [Doctest]change distance docstring, test=docs_preview

* [Change]add ref

* [Change]patch xdoctest debug

* rename weight_only/llm.int8

* add clip_grad_value_ api

* add test for ClipGradByValue

* typo fix

* refine and modify clip_grad_norm_

* no_grad

* clip_

* remove g=p.grad

* bug: AssertionError: When Variable is used as the condition of if/while , Variable can only contain one element.

* add linear_compress API

* COPY FROM issue of paddle nn

* 移除额外添加的示例，并修改Examplex->Examples

* remove example name of transformer

* add pos_weight in kernel

* fix unittest

* fix xpu

* fix bce unittest, change infermeta order

* move batch_norm prim test to op_test

* fix optest bug

* add test to cmake

* add cinn test case

* fix batch_norm prim grad bf16

* fix code

* add cuda check

* fix batch_norm bfloat16

* fix cpu bfloat16 bug

* skip non-bfloat16-supported platform

* fix code

* fix cinn rtol and atol in bfloat16

* fix name

* fix config

* bump ruff to 0.0.271 and update config

* exclude third_party

* bump ruff to 0.0.272

* refine config

* [dygraph]unify _non_static_mode() in_dygraph_mode() and in_dynamic_mode()

* [dygraph]unify _non_static_mode() in_dygraph_mode() and in_dynamic_mode()

* [dygraph]unify _non_static_mode() in_dygraph_mode() and in_dynamic_mode()

* [dygraph]unify _non_static_mode() in_dygraph_mode() and in_dynamic_mode()

* [dygraph]unify _non_static_mode() in_dygraph_mode() and in_dynamic_mode()

* [dygraph]unify _non_static_mode() in_dygraph_mode() and in_dynamic_mode()

* fixed cyclic reference that caused patial import

* fixed bad change

* fix bad import

* fix bad import

* fix bad import

* fix ut failed caused by change in_dynamic_mode

* fix ut failed caused by change in_dynamic_mode

* fixed usage of in_dynamic_mode() or in_dygraph_mode()

* revert python3 to python in .pre-commit-config.yaml

* fix merge conflicts

* fix parameter not passed

* fix repr

* simplify layer_norm_op.cc

* support auto generate for op layer_norm

* update unittest for composite_layer_norm

* remove layer_norm_op.cc from scripts

* replace layer_norm_op with generated_op

* add get_expected_kernel for layer_norm

* update cmake kernel register function for layer_norm_mkldnn_op

* support fp16 for maxout op

* format code

* change api

* add test for static float16

* format code

* formatting code

* atol alignment

* experiment—1

* experiment-2

* experiment-3

* format code

* fluid clean: remove print/switch from fluid to static

* remove Switch in static.__init__

* fix conflicts.

* replace Switch by case.

* fix piecewise_lr decay.

* fix typo

* fix conflicts.

* fix lr dtype

* keep Switch in paddle.static.nn.control_flow and fix piecewise_lr.

* fix conflicts.

* keep Switch in the fluid.

* fix Switch doc

* fix example in Switch doc

* fix Switch doc.

* fix static/__init__.