* static graph autogen code support for softmax op

* bug fixed

* fix PR-CI-Windows error

* fix CI error

* bug fixed

* fix conflicts

* remove _in_eager_mode_

* remove _in_eager_mode_

* Allow to switch whether to use promote strategy to choose kernel for O2 training.

* Fix comparing error and add unittest.

* fix embedding model weight type mismatch error

* Update fp16_utils.py

---------
Co-authored-by: NZhang Ting <zhangting_2017@163.com>

[inference Zero-Dim][trt] Add Zero-Dim tensor support for clip, cast, flatten_contiguous_range (#53769)

* [inference Zero-Dim][trt]clip,cast,flatten_contiguous_range trt op converter support zero dim

* add rules

* modify no kernel yaml parse

* success op generate

* success test_silu_double

* modify bug

* modify static error

* modify silu_grad input

* modify kernel signature

* modify kernel signature

* code style

* code style

* review

* delete opinfo modify

* relocate regularizer.py

* fix bug

* fix bug

* fix bug

* relocate the import

* replace _regularization_coeff with coeff

* remove the L1DecayRegularizer and L2DecayRegularizer

* modify concat_grad add sum comp rule

* delete default mul_double_grad

* delete high grad test

* recover yaml

* modify yaml

* recover add_double_grad prim

* test(prim-cinn): split test_resnet and test_bert into three tests

* test(prim-cinn): fix cmake file to run prim test in CINN-CI

* softmax support

* support stack

* sequence_mask functionalization

* fix sequence_mask test

This reverts commit 45ce0ad5.

* [Dy2St]Fix x grad names when high order gradient

* Polish error msg

* Add inputs var to backward in dy2st

* Fix error

* Get grad names for backward API

* Fix save load

* Polish code

* Add ut

* [prim] fix not support optional grad bugs in higher order autodiff

* [prim] remove duplicate fill_any_like caused by infershape_for_composite

* fix _strip_grad_suffix_ bugs in higher-order autodiff

* [prim] create output for test_static_prim.cc

---------
Co-authored-by: N0x45f <wangzhen45@baidu.com>

* support_act
* delete_silu

* prelu op trt converter support zero dim

* move DataLoader to paddle.io. test=develop

* add pad op for xpu

* add pad op for xpu

* add pad op for xpu

* modify concat_grad add sum comp rule

* delete default mul_double_grad

* delete high grad test

* recover yaml

* modify yaml

* [test]mv fluid controlflow detection dlnne tensorrt tests to tests

* [test]clean dlnne

* [test] fix test_tensorrt_engine_op

* [test] try fix path error

* [test] RollBACK test_tensorrt_engine_op

* [test] RollBACK test_tensorrt_engine_op

* [test]add todo

* Empty-Commit; test=document_fix

* support_expand_mish

* add_depthwise_conv2d_transpose

* Update test_depthwise_conv2d_transpose_op_xpu.py

删除print语句

* Fix div error when dtype is int64 in static mode

* Fix out dtype

* fix as review, add fp16 conv2d_transpose

* fix unittest of bn and reduce_mean

* fix bn unittest

* fix ci

* fix ci

[CINN]Adjust Bert unittest loss ground truth, see: PaddlePaddle/CINN#1357

* [Zero-Dim] add 0D Tensor UT case for XPU

* fix comment

* remove some unnecessary UT

* bind sparse_coo_tensor, reduce_max/max_int32, range/arange_int32, equal_bool, scatter_grad_float32, nearest_interp_int64 kernels

* add more unit tests; modify compilation logic of xpu sparse kernels

* delete delete_fill_constant_op_pass and unsqueeze2_eltwise_fuse_pass

* 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