* Chage `IndicateDataType` to `GetKernelType`. Make it easier to
  understand.
* Change `OpKernelKey` to `OpKernelType`
* Make operator developers can customize which kernel the operator will
  use in runtime.

* small fix

* fix comment

* address comment

* small fix

* Adding design doc for model average (now called parameter_average)

* Updating title

* Updating image tag

* Updating review comments

* Init

* Stash

* Polish SaveLoadOp

* Fix CI

* Polish code

* Save GPU Tensor

* Stash

* Fix CI

* Removing survey out of the design doc and fixing typos
* Fix Typos

* Extract apply_backward_pass to backward.py

Rename apply_backward_pass to append_backward_ops

* Fix CI

* Update design doc

* "add model format design doc"

* "add restore function"

* "add parse protobuf"

* "move necessary information to saver.proto"

* "format code"

* "add gpu option"

* "add lod info"

* "add saveop python test wrapper"

* "checkpoint reuse save operator"

* "rewrite model format design doc"

* "async support needed"

* "fix run once"

* "fix doc based on comments"

* "refine based on comments"

* "fix based comments"

* "remove persistable flag from framework.proto"

* "add IndicateDataType to restore op"

* "add save test"

* "modify save restore code"

* "modified the restore logic"

* rm checkpoint_op.cc

* rm test_checkpoint_op.py

* "get inputs outputs name from execution context"

* Saving each variable to a independent file

* Fix bugs

* Rewrite save_restore_op_test with new Python framework

* Move `SaveOp` and `RestoreOp` from OpWithKernel to OpBase

* Refine unit test of SaveOp and RestoreOp

* fix compile errorwq

* update cluster train v2 doc

* WIP cluster train doc

* update

* cluster train doc

* add TOC for en doc

* fix sphix build issue

* fix error links

* fix link errors

* fix image link

* polish cluster train docs

* update general distributed training document

* fix sphinx compile error

* fix doc image error

* Create prune.md

* modification based on comment

* remove insertion

* rename id to block_id

* Update prune.md

* formatting

* Add initail design doc for regularization

* Updating image links

* Commiting the images for the equations

* Adding computation graph images

* Adding section on computation graph

* Update design doc for Python Layer

* Update document

* Design doc of SelectedRows

* Follow comments

* Update protobuf message

* Follow comments, seperate LoDTensorDesc and SelectedRows Desc

* init optimizer design

* fix index

* optimize the interface

* add a link to python_api.md

* optimize the code of Optimizer