* add enable_imperative, disable_imperative; test=develop

* add unitest; test=develop

* polish example; test=develop

* Remove the NGraph engine from PDPD repository
1. Each operator was removed from the operator's directory
2. Each test was removed from the unittest directory
3. The parallel executor support was removed from the PDPD
4. The CMake file was removed from the PDPD
5. The NG flags were removed from the repository
test=develop

* Remove ngraph from:
1. Cmake file
2. Python file
test=develop

* fix init_gflags with 'python -c', test=develop

* add test, test=develop

* use sys.executable instead of python, test=develop

* keep dummy, test=develop

* add recorded cuda memory apis, fix typo, test=develop

* add more ut, test=develop

* follow comments, test=develop

* fix py35 incompatible issues, test=develop

* functional dygraph enable/disable
test=develop

* use context manager instead
test=develop

* refine sample code
test=develop

* rename api & expose to fluid
test=develop

* fix sample code
test=develop

* add half_async in the communicator
* fix DistributedStrategy

* add some debug flags to auto growth allocator, test=develop

* add comments about auto growth, test=develop

* test=develop, fix docker with paddle nccl problem

* don't expose numerous Tensor.set(), test=develop

* fix condition, test=develop

* fix float16 bug, test=develop

* feed should be Tensor or np.array, not Variable or number, test=develop

* use forcecast to copy numpy slice to new array, test=develop

* remove float16-uint16 hacking, test=develop

* add variable method to varbase and refactor to_variable to support return varbase

* support kwargs in varbase constructor

* add VarBase constructor to support default python args

* refine varbase initial method

* reset branch

* fix ut for change VarBase error info to PaddleEnforce

* cherry is parameter change before

* overload isinstance to replace too many change of is_variable

* rm useless files

* rm useless code merged by git

* test=develop, fix some ut failed error

* test=develop, fix test_graph_wrapper

* add some tests, test=develop

* refine __getitem__, test=develop

* add tests, test=develop

* fix err_msg, test=develop

* add unused input vars check for OpWithKernel, test=develop

* remove unused vars in some ops, test=develop

* fix batch_norm, test=develop

* add white list, test=develop

* add CI check for white list, test=develop

* :ove white list to c++, test=develop

* solve failure of CI, test=develop

* add unittest for unused_var_check, test=develop

* refine code, enable check in operator_test, test=develop

* skip mkldnn, test=develop

* extend white list, test=develop

* refine condition of mkldnn, test=develop

* fix paddle_build, test=develop

* follow comments, test=develop

* fix GetExpectedKernelType

* add wiki ref to err_msg, test=develop

* follow comment, test=develop

* use system allocator in unittests, test=develop

* fix op bugs, test=develop

* fix tensor copy bug when src and dst are the same, test=develop

* fix persistable;

* fix save load bugs; test=develop

* fix bug; test=develop

* add example for new io api; test=develop

* addd example; test=develop

* test=develop, communicator_is_sgd_optimizer flags

* add new save load interface; test=develop

* add new save interface; test=develop

* add save load interface ;

* fix save load error;

* fix dygraph set dict bug;

* add save load unit test; test=develop

* fix test_imperative_optimizer bug; test=develop

* fix unitest optimizer bug; test=develop

* fix code coverage; test=develop

* fix converage; test=develop

* add document for apis; test=develop

* fix unitest error; test=develop

* fix save load unit test error; test=develop

* fix error message; test=develop

* change set_parameter set_optimizer to save_dygraph; test=develop

* add load_graph check; test=develop

* fix api spec; test=develop

Heartbeat for distributed async training.

* refector geo sgd & communicator

The new "fluid.data" changes old "fluid.layers.data":

1. Add shape and dtype check.
2. Remove "append_batch_size" parameter. We won't offer this in the new data layer because other deep learning platforms don't have this kind of data layer pre-processing. It may confuse users.
3. Remove "stop gradient" parameter because the data layer doesn't do back-propagation

TODO：
Now data layer feeded by executor is checked, will we want to check the feed data of readers in the future?

* Remove constraint that last dimension is forced to be 1 by add
lookup_table_v2 test=develop

* modify into PADDLE_ENFORCE_CUDA_SUCCESS test=develop

* Revert "modify into PADDLE_ENFORCE_CUDA_SUCCESS test=develop"

This reverts commit 8a960bfc61e51aa27c3c529df8fb90b93ebd19f9.

* move api into fluid.embedding test=develop

* fix example code test=develop

* move one_hot into fluid.one_hot

* modify api.spec test=develop

* fix loss shape test=develop

* Add RecordHistoryLocalExecScopes
test=develop

* rpc retry for asycsend/get/prefetch

* test=develop, change retry vlog level to 3

* test=develop, set default grpc_retry_times is 3

* add one_hot_v2_op to remove last_dims==1 test=develop

* add api unittest code for CI_Coverage test=develop

* improve CI_Coverage rate by adding test_with_depth test=develop

TemporaryAllocator is a singleton used for allocating memory for Cudnn. Since it is a singleton, we can delete it for better performance in memory.

We replace TemporaryAllocator by CUDADeviceContextAllocator and CUDADeviceContextAllocation, which uses stream callback to delete the memory allocated for the stream to avoid singleton.

Also added data_feed_proto to operator to fix CI in CPU compilation

test=develop

* remove legacy memory optimization codes, test=develop

* follow huihuang's comments,test=develop

* follow luotao's comments, test=develop

test=develop

* optimize communicator merge sparse gradient test=develop

* revert multithread selected rows merge add test=develop

* follow comment test=develop

* refine core cmake warning and print more info

test=develop

* fix comments

test=develop

* fix dygraph mem leak, test=develop

* polish msg, test=develop

* Update backward.py:
     - If there is no input grad var in all outputs of previous ops, do not append this op into graph.
     - Only apply this stragety when double backward.
* Update some double backward op.
* Update sum_op to judge whether a tensor is empty by numel or IsInitialized().

* add gradient clip in minimize; test=develop

* fix bug; test=develop

* fix format; test=develop

* move new grad clip to dygraph/grad_clip.py; test=develop

* fix lr decay and grad clip test; test=develop

* seperate dygraph grad clip; test=develop

* fix grad clip test; develop

* fix api spec bug; test=develop

* add blank line, test=develop,test=document_preview

to fix format problem

* optimize communicator flag

* change flags in init py test=develop

* remove unused FLAGS_warpctc_dir

test=develop

* remove FLAGS_warpctc_dir

test=develop

* remove async executor python api
test=develop

* remove test_async_executor.py
add executor train_from_dataset demo
test=develop

* fix import bug
test=develop