[CodeStyle] replace `assert np.allclose` with `np.testing.assert_allclose` and `assert np.array_equal` with `np.testing.assert_array_equal` (#55385)

* [Paddle-TRT] add flip op

* Paddle-TRT support float64  in/out type, support fill_any_like_op in int64

* onednn: remove fc+eltwiseadd fusion pass
* onednn: remove post-sum fusion in fc kernel
* onednn: tests: make unfused add run into f32

* Fix TRT multihead matmul UT failure

* delete relu6_raw

* fix codestyle

* Update test_mkldnn_matmul_activation_fuse_pass.py

* fix

* Update backward.yaml

* Update ops.yaml

* Update backward.yaml

* unary bitwise_not adapter tensorRT8.6 in Paddle-TensorRT
* Update test_trt_inspector.py
* test_trt_convert_conv2d_transpose
* Update test_trt_convert_conv2d_transpose.py

* support_conv2d

* remove comment

* check code style

* add former Test

* check code style

* add unittest

* fix log

* change unittest

---------
Co-authored-by: zhoutianzi666 <17801055074@163.com>

* [Paddle-TRT] add assign op

* rewrite identity_op_clean_pass

* fix

* adjust identity_op_clean_pass order in gpu passes

* fix ut

* add einsum layer

* Enhance the shape check of trt_embedding_eltwise_layernorm_fuse_pass, embedding_eltwise_layernorm_fuse_pass

* delete swish_raw

* fix

* Update activation_kernel.cc

* fix

* set_value update

* support ValueTensor's rank != Input'rank & update topk

* update range to avoid coredump

* fix addShape error

* Dims definition differ between 7.2 and 8.0+

* Update test_trt_convert_top_k_v2.py

* update top_k

* Update test_trt_convert_top_k_v2.py

* 0-dims support cumsum and bitwise_not
* Update cumsum_op.cc
* Update test_trt_convert_bitwise_not.py
---------
Co-authored-by: NZhang Jun <ewalker@live.cn>

* modify tensorrt ci timeout

* activation ci bug fix

* comment out  int8 mode test_trt_dynamic_shape_groupnorm

* layer_norm op with dynamic shape support INormalizationLayer in TRT8.6

* Using trt layer to make layers_norm op in lower than trt8.6
layer_norm op with dynamic shape support INormalizationLayer in TRT8.6

* remove redunant

* unary bitwise_not adapter tensorRT8.6 in Paddle-TensorRT

* one_hot_op ci fix

* Update test_trt_inspector.py

delete log

* update roll convert

* fix fp16 io

* disable precision test

* add tranfer_elim
* transfer layout elimination

* [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

[Inference] add config.enable_low_precision_io api and remove rely on AnalysisConfig::Precison in trt (#52485)