1. Fix kernel
2. Add more test case

2. Remove num_seq arguments.
3. Refine CUDA kernel of ScaleLoDTensorFunctor.
4. Change max_relative_error of gradient unitest to 0.007

add wrapper for element-wise math operator.

* update code comments.

* update the comments.

* follow comments.

add how to update whl in build_from_source.md

enbale auto set env of V1 on Mac

fix display error of C-API doc.

* add GetLoD for debug

* add LoDToString

* optimize if

* typo

* add lod_tensor to operator's dependency

Adding the rst file to render IO under Fluid in paddlepaddle.org

add doc for how to use C-API.

assign_value operator

* add cluster training bencharmk design

* update by comment

* update by comment

Fix the documentation for elementwise op in fluid layers

Add book demo distributed fit a line

Book demo understand sentiment distributed version

Add distributed label semantic role book chapter

* Import content from the buildtools repo

* Renmae Dockerfile-x86_64 into Dockerfile.x64

* yapf

* Remove subdirectory root

We need this operator to assign value to a tensor and the values are stored in the program so that they can be used independent of python.

Feature/parallel for unittest

Feature/sync wait