* extend matmul op to support multiple head multiplication

With the support of multiple head, the multiplication of two big matrixes is
split into multiplication of several (head_number) small matrixes. e.g. if
Mat A is [3, 24] and Mat B is [24, 4], when multiple A and B with head_number
as 4, Mat A will be split as 4 matrix of [3, 6] and Mat B will be 4 matrix of
[6, 4]. The result of final matrix will be 4 matrix of [3, 4], i.e. [3, 16].

* split different comm method for mnist distributed training

* feature/auto_growth_allocator, test=develop

* add unittest of AlignedAllocator, test=develop

* try to turn on auto_growth to test on CI, test=develop

* fix segmentation fault in mixed_vector.h, test=develop

* add unittests, test=develop

test=develop

2）add pip install coverage in Dockerfile.tmp
test=develop

* feature/buffer_shared_inplace, test=develop

* refine code, test=develop

* fix elementwise_add op cpu inplace and sum inplace bug, test=develop

* add unittest and debug log, test=develop

* fix parallel_executor scope bug, polish code, test=develop

* fix sum op, activation op, single_in_place_inference bug, test=develop

* remove kLocalExecScopeName, test=develop

* fix unittest,test=develop

* fix out_var first version bug, test=develop

* follow comments,test=develop

* add WITH_COVERAGE option, default OFF

test=develop

* add coverage for python sdk

test=develop

* fix code style

* fix COVERAGE_FILE path

test=develop

* remove coverage package

test=develop

* test = develop, run coverage as module

* fix prepare context redundant code problem, optimize executor by caching create_varaiables
test=develop

* supports collective training in executor

* make fetch_list runable with variables, add more unittest for use_program_cache
test=develop

* fix comment
test=develop

* use unique name for nccl_id

* supports output to stream in program_to_code

* insert sync_comm_stream before regularization; add skip_op_callstack capability in program_to_code

* set op role in collective training

* add collective op role

* remove orig file

* add build optimizer by strategy

* add collective strategy

* refine collective strategy

* add multi-process role maker

* refine strategy building factory so that we can easily plugin more strategy

* scale loss grad in collective sgd transpiler

* add support for distributed fc

* code format

* revert some features for dist fc

* add support for distributed fc training

* fix prepare context redundant code problem, optimize executor by caching create_varaiables
test=develop

* supports collective training in executor

* make fetch_list runable with variables, add more unittest for use_program_cache
test=develop

* use unique name for nccl_id

* supports output to stream in program_to_code

* insert sync_comm_stream before regularization; add skip_op_callstack capability in program_to_code

* set op role in collective training

* add collective op role

* fix comment
test=develop

* remove orig file

* add build optimizer by strategy

* add collective strategy

* refine collective strategy

* add multi-process role maker

* refine strategy building factory so that we can easily plugin more strategy

* scale loss grad in collective sgd transpiler

* add support for distributed fc

* code format

* revert some features for dist fc

* add support for distributed fc training

* test=develop
add collective op unittest standard

* test=develop
remove the test_collective directory

* test=develop
remove the test_collective directory

* remove slicegather test

* code format for reducescatter

* update attr of shard_index_op

* Modify macro nccl_helper

* remove test without distribute

* macro collective_helper

* marcro update

* test=develop
update support python3.5

* test=develop change gpu memory use to 0.1 when test

* test=develop
update ut equal func

* test=develop
set flags to 1.5

* test=develop fix pickle dumple  py35

* test=develop
fix divide in slice and add sync_comm_stream
update atol and rtol to 1e-05
rm shard_index op and test
modify read input from file to read from memory
remove origin_program in framework and add i/o in c_sync_calc_stream

* test=develop update unittest sync operator I/O

* sequnce mask support max length tensor input; test=develop

* add rnn_impl.py; test=develop

* add basic gru lstm unittest; test=develop

* fix api spec; test=develop

* fix sequence_mask op bug;
test=develop
test=document_preview

* change +-*x to elmentwise_op; test=develop

* add mkl flag; test=develop

* fix rnn impl bug; test=develop

* update api spec; test=develop

* fix doc bug; test=develop

* fix lstm bugs; test=develop

* add more print function for timeout issue, make timeout value larger

* test=develop, fix test_imperative_transformer and ocr

* test=develop, remove ocr recovery part

* test=develop, add add_multi_gpu_install_check

* test=develop, refine warning doc

* test=develop, refine warning doc

* test=develop, refine warning doc

* test=develop, support multi cpu

* remove nccl dep when the number of GPU is 1
test=develop

* fix save/load in Fleet
* add UT framework of Fleet

test=develop

Split the unittest test_dist_mmist into multiple unittests (test_dist_mnist, test_dist_mnist_nccl and test_dist_mnist_lars) to avoid timeout (#17707)

* test=develop, add ocr in dygraph test

* test=develop, add cudnn determinist

* test=develop, remove useless code

* test=develop, fix cmake error

test=develop

* test=develop, add gradient sort backward strategy

* test=develop, fix test by add FLAGS_cudnn_deterministic on new tests

* remove unused FLAGS_warpctc_dir

test=develop

* remove FLAGS_warpctc_dir

test=develop

* test=develop

* test=deelop

implement dygraph.parallel.DataParallel to hook reduce op.

Support backward of backward for Relu and add a new gradient checker by comparing theoretical and numerical Jacobian. (#16862)

* Support backward of backward and a new gradient checker
* Rename decorators.py to decorator_helper.py, since Python on Windows CI has decorators package.

1. Add ReluDoubleGradMaker when register relu_grad.
2. Add a new gradient checker by comparing theoretical and numerical Jacobian.  Check double gradients by double_grad_check.

* move gc test to op_test
test=develop

* Revert "move gc test to op_test"

This reverts commit cf15da65c38f57c91f53b3d8b3c2365d4aa86016.

* enable gc test in some ops
test=develop

* add parallel build script to ci test=develop
* 1. classify the test case as single card/two cards/multiple cards type
   2. run test case according to the run type

* enable recurrent op test=develop

test=develop