We put forward Op's inputs, outputs and output gradients into Grad
Op's inputs, and put forward Op's input gradients into Grad Op's output.
So Grad Op's `in_out_idx`, `input_format` and 'output format' need to be
rebuilt during Op creating.

* Fc operator is a grouped operator, which combined by may internal
  operators.
* InferShape & Run a FC operator in Python.

* Not use __CUDACC__ macro, because malloc in GPU memory do not have to
  used in .cu file.
* Also, `PADDLE_ENFORCE` logic is wrong.

* `in_out_idxs_` shares between all operator instance in same type
  of operator.

Net::DebugString() print what it is and invokes inner operators
DebugString() with a prefix '\t'.

* There is a merge conflict when merge PR #2914
* Develop and PR #2914 both add `DDim::size` method, but did not
  triger git merge conflict while merge.

* Follow review comments to seperate Tensor Numpy interactive methods in
  tensor.h.
* Simplify logic for `CastToPyBufferImpl`, make it as one struct and in
  details namespace.
* Remove `Scope` expose in Python, since it currently is useless.
* Remove some debug functions.

Basically following
http://pybind11.readthedocs.io/en/stable/advanced/pycpp/numpy.html

* Use buffer protocol to return a view of Tensor. It can be cast to
  numpy array in Python.
* Set a numpy array to a tensor.

* Also fix unit test

* add inputs

* add ut for multiple inputs

* fix AddToLayer

* op_desc -> op_proto

* CreateArgumentOffsetMap -> CreateInOutOffsetMap

* move CreateInOutOffsetMap from OperatorBase to op registry

* arg_idxs_ -> in_out_idxs_

* Implement InferShape and register them, give a stub Kernel method
  by LOG(INFO)

* add NDEBUG switch to PADDLE_ENFORCE

* using etcd as fault tolerant training

* update

* workable version, ft not tested

* small fix

* update

* remove TODO