* add stat tool

* add roll and roll_grad kernels and strided_slice and strided_slice_grad kernels, test=kunlun

* add masked_selected_grad kernel,test=kunlun

* Remove API: label_smooth (replace with paddle.nn.functional.label_smooth)
	Replace the paddle.fluid.layers.label_smooth with the
paddle.nn.functional.label_smooth

* modify the call of label_smooth from old style to new style

* fluid clean move out LayerList, ParameterList, Sequential from fluid

* copy Sequential from fluid to paddle.nn

* replace fluid LayerList,ParamterList, Sequential in unit test file

* add omit unit test file use paddle.nn.Sequential

* remove numpy and use paddle.uniform instead

* first commit

* 2nd commit

* remove transpose

* codestyle check

* modify CI_STATIC

* modify CI_STATIC

* modify enable static()

* remove unused import

* fix conflict about stack

* fix conflict about stack

* fix conflict about stack

* endless conflict

remove dice_loss which is not used in paddle 2.0

* clear fluid apis: center_loss, bpr_loss, edit_distance, hsigmoid, sampled_softmax_with_cross_entropy, rank_loss, margin_rank_loss,  sigmoid_cross_entropy_with_logits, huber_loss

* fix python/paddle/fluid/layers/loss.py

* fix test_layers.py

* fix CI bug

* fix nn.py

* [remove fluid] under fleet meta_optimizers_wz

* [remove fluid] under fleet meta_optimizers_wz

* update

* [remove fluid] under fleet meta_optimizers_wz

* [remove fluid] under fleet meta_optimizers_wz

* [remove fluid] under fleet meta_optimizers_wz

* [remove fluid] under fleet meta_optimizers_wz

* [remove fluid] under fleet meta_optimizers_wz

* [remove fluid] under fleet meta_optimizers_wz

* [remove fluid] under fleet meta_optimizers_wz

* [remove fluid] under fleet meta_optimizers_wz

* [remove fluid] under fleet meta_optimizers_wz

* [remove fluid] under fleet meta_optimizers_wz

* [remove fluid] under fleet meta_optimizers_wz

* [remove fluid] under fleet meta_optimizers_wz

* [remove fluid] under fleet meta_optimizers_wz

* [remove fluid] under fleet meta_optimizers_wz

* [remove fluid] under fleet meta_optimizers_wz

* Skip tests that use fused_ops on H100

* Add error message to FusedOps on H100

* remove hard_sigmoid in nn.py under fluid

* fix hardsigmoid test case

* fix hardsigmoid test case

* remove reshape in nn.py under fluid

* remove reshape reference

* fix test case

* fix test case in distribution.uniform

* remove fluid reshape reference

[CodeStyle][py36-][E722] remove import handling for collections.abc in different python versions (#48165)

paddle.scatter_nd_add

* remove apis in fluid.ops

* fix test_activation_nn_grad

* fix circle import error

* fix ops

* fix cos

* fix divide not inplace

* remove lazy-import part

remove crop which is not used in Paddle 2.0

* remove relu6 test case under fluid

* fix relu6 test case in mkldnn_elt_act_fuse_pass

replace paddle.fluid.layers.selu with paddle.nn.functional.selu

* Remove API: gather
	replace the paddle.fluid.layers.gather with paddle.gather

* modify the call of gather from old style to new style

* remove range

* [CustomPlace] fix amp

* [CustomPlace] fix amp

* fix ut because of too long time matmul fp16

remove pad_constant_like which is not used in paddle 2.0

* [AutoParallel] selective recompute

* add cmakelist