* Refine the tester for MixedRTPredictor.
test=develop

* Enable the profiler in TensorRT engine.

* Support the use of combined inference model in TensorRT unittest, and print the shape of feed targets.

Softmax for Inference is enabled when ON_INFER is set

add plugin support for paddle-trt

Refine the pass builder and buildstrategy

test=develop

Fix compiling with cuDNN v5.

test=develop

Adding fused operator for ngraph 

* Split cichecks

test=develop

* feat(Anakin): expose anakin options to paddle cmake option

Expose ANAKIN_BUILD_FAT_BIN, ANAKIN_BUILD_CROSS_PLANTFORM to Paddle cmake option

test=develop

test=develop

fix prelu

merge develop and fix conflicts

* human readable memory warns test=develop

* update test=develop

* refine test=develop

* fix build test=develop

Refine lrn cpu forward

Lookup sparse table add test mode

sgd_op optimize selected rows do not enforce id < height

Fix num_threads settings in fast_pe

Revert "Softmax op optimization for inference "

test=develop

add self to author

test=develop

test=develop

Softmax op optimization for inference 

test=develop

test=develop

test=develop

split async executor into executor_thread_worker and async_executor, refactor pybind, add datafeed and corresponding proto

test=develop

test=develop

test=develop

test=develop

test=develop