add fp64 support in sequence_pool, test=develop

* add more vision transfrom apis

* fix best_fit_allocator_test on windows, test=develop

* enable best_fit_allocator_test and test_math_op_patch_var_base, test=develop

* fix dataset py3 error
* test=develop

* add ci check for changing op-related api without core.ops, test=develop

* generate api_source_md5 file when build, test=develop

* add failed example, test=develop

* add failed example, test=develop

* handle exception, test=develop

* add op multiply, delete op elementwise_mul from math.py. test=develop,test=document_fix

* bug fix. test=develop,test=document_fix

* bug fix. test=develop,test=document_fix

* bug fix. test=develop,test=document_fix

* bug fix. test=develop,test=document_fix

* add unittest for multiply op. test=develop.

* fix code style. test=develop

* use random input. test=develop

* add test error case for static computation graph. test=develop

* add np.random.seed(7)

* increase input ndarray size. test=develop

* change float32 to float64. test=develop

Modify the print in Dy2stat for dynamic type. Unit test is covered in old test_print.py

* add reinforcement learning model test=develop

* align backward test=develop

* add gym in paddle_build.sh test=develop

* rm pip install in script test=develop

* refine paddle_build.sh test=develop

* fix sed error in macOS test=develop

* polish code test=develop

* fix scope problem

* refine code by reviewer comment

just skip case of distribute on windows

* update cross op parameters

windows CI scripts for xly

We found that the reason of multiple return error in SimNet is that I wrote wrong task_mode. If we set task_mode as "pairwise" correctly, which is a format of the model input data, the multiple return won't have problem in the unit test. In this PR we corrected the task_mode and enable multiple return in SimNet unit test.

refactor fleet api under paddle.fleet
update DistributedStrategy

* add reinforcement learning model test=develop

* align backward test=develop

* add gym in paddle_build.sh test=develop

* rm pip install in script test=develop

* refine paddle_build.sh test=develop

* fix sed error in macOS test=develop

* polish code test=develop

* refine infer ir pass tests, test=develop

* commit to change copy_from_cpu input to np array, test=develop

Enable multiple "return" statement in MNIST as Unit test in dy2stat.

* fix test_activation_op, test=develop

* remove paddle.mul related unittest, test=develop

* delete mul op. test=develop,test=document_fix

* bug fix. test=develop,test=document_fix

Add Support for SelectedRows for Transpose OP and Fix a Bug That SelectedRows Cannot be Supported in SimNet (#25536)

This PR fixes a bug that SelectedRows cannot be supported in SimNet. The reason of this bug is that dygraph basic_engine didn't copy var's type when the var needs to be accumulated during backward. So when a var is SelectedRows and needs to be accumulated, like SimNet which calls net for two times, the var's type will be changed to default LoDTensor thus bug happens. To fix it, we just also copy the type.

Without this PR, the accumulated SelectedRows parameters in dygraph will be changed into LoDTensor. So when we fixed the bug of supporting SelectedRows in SimNet, we found `test_imperative_lod_tensor_to_selected_rows` failed and threw the error that SelectedRows was not supported for Transpose OP. To fix it, too, this PR also added support for SelectedRows for Transpose OP.

* remove op call stack test=develop

* fix typo test=develop

* fix doc of logical_* op.

* fix doc of op pow.

* fix comment syntax error9D

* fix operator reciprocal demo.

* fix logical_* ops' doc. test=develop,test=document_fix

* bug fix. test=develop,test=document_fix

* bug fix. test=develop,test=document_fix

* bug fix. test=develop,test=document_fix

* bug fix. test=develop,test=document_fix

add try..catch...

* add bilateral slice op

* randint API: remove out, devive, stop_gradient; add name; test=develop

* test=develop

* test=develop

* test=develop

According to paddle 2.0 standard
1, change flip api attr name 'dim' to 'axis'.
2, support empty axis
3, change example code to imperative mode.