Add WMT16 dataset.

* add memory optimization transpiler demo

* add memory benchmark compile option

* add gflags instead of macro

* refine code

Gradient clip by global norm

Add math operator patches

Fix priorbox layer when multiple values given in min_size

Add math function for sampling integers

refine openblas install lib when CAPI

[WIP] Split SelectedRows to multiple pservers

* init complete data layout transform

* can compile

* test passed

* optimize code

* fix while_grad_op first step loss lod problem

* optimize in out ptr for transform

* add check

* update copyright

* clean code

* add NeedTransformLayout

* add comment

* change the interface of data_type_transform

* init data_type_transform_test

* complete data_type_transform_test

* add TransDataType to data_transform

* Add fluid_compiler.md

* Paragraphing

* "fix decode bug"

* "follow commnet"

* "fix error"

* "fix hook bug"

* fix based comment

* fix copyright

* fix based on comment

Enhance matmul_op to support 4-D inputs.

remove unused vars which block the build

Recv OP: use BlockDesc* instread of ProgramDesc proto as Attribute

Fix the description of default scale attribute for scale op.

Add distribution implement of image classification

Add python wrapper for sequence reshape operator.

Add sequence reshape operator