* Implementing the DecayedAdagrad optimizer step operator

* implementing DecayedAdagrad operator

* remove file

* small fix

Feature/fix ci

* add cudnn_conv_op

* WIP

* update

* update

* fix grad check

* use platform::memory

* add support group for cudnn

* update

* follow comments

* fix onlycpu build

* update cuda define

* follow comments

* follow comments

* merge with updates

* fix compile error

* follow comments

* follow comments

ImportError: /usr/local/lib/python2.7/dist-packages/py_paddle/_swig_paddle.so: cannot allocate memory in static TLS block

* Adding thresholded_relu op
* Adding test for thresholded relu op

* init

* unify CopyFrom interface

* fix gpu build error

* fix bug in tensor_py.h

* refine code comments and add TODO list

* fix conflicts in FeedOp and FetchOp

Explose check_attr to Python

expose AppendBackward of ProgramDesc to python

Update annotations of layers.py.

Add the sequence_concat operator.

fix executor_test build failed

Add margin rank loss operator

use protobuf_lite druning inference

Fix pool op doc and class declarations

fix Compile error [fatal error: boost/range/adaptor/reversed.hpp]

Simplify backward when inserting a sum operator to accumulate all duplicated variables