* fix c_split bug

* fix utest

* add c_embedding for tensorparallel

* fix hang of hybrid parallel

* fix new_group for hang problem

* Dygraph Recompute support AMP

* dygraph recompute: update unitest

* update mp

* fix spawn default nprocs get error

* polish error message

* support ssd in PsCore

* remove log

* remove bz2

* defalut value

* code style

* parse table class

* code style

* add define

* add async save for sparse table
* add load profiler for sparse table
* add load info for sparse table

* fix pipeline

* fix mp pp dp

* fix utest of hybrid parallel

* add utest for tuple

* fix precision of mp

* fix bug of seed

* fix dp

* print group

* fix cpp lint
* fix save/load with unexpected value
* fix save and user interface

* fix find_unused_parameters default value

* add trainprofiler for heterps in oneps; test=develop

* add set_use_ps_gpu; test=develop

* add raw program, test=develop

* update, test=develop

avoid polluting logging's root logger

* Revert "Revert "[PsCore] optimize performance of large kv (#32535)" (#32599)"

This reverts commit 809ac036.

* brpc dep

* add alltoall api, test=develop

* fixed docs.

* Fixed docs. test=document_fix

code bak.

fixed docs. test=document_fix

* Revert to previous version of python/paddle/fluid/backward.py

* fixed bugs.

* test=document_fix. Fixed examples.

* Support list and tuple for parameters of layer_norm, multiprocess_reader, DatasetFolder and ImageFolder.

* add unittest for layer_norm.

* add require gpu for example.

This reverts commit 4b7242b0.

* fix amp bug

* fix name of wordsize

* add sendrecv, test=develop

* optimize pull sparse

* optimize pull sparse

* change macro

* format

* fix model parallel

* rm parallel_help.py

* add embedding

* add pp dygraph, test=develop

* Dygraph reocmpute

* unitest for Dygraph reocmpute

* dy recompute remove unitest for win and mac

* add pipeline layer

* update

solve hccl communicate conflict (#32447)