clean up channel

test=release/1.0.0

(cherry picked from commit 425a8821)

* refine reduce by cub
* optimize KernelDepthwiseConvFilterGrad
* optimize depthwise conv and reduce mean and reduce sum
* fix bug: dilation
* cuda arch and cuda 8 compatible

test=release/1.0.0

* refine seq_softmax_op

* fix seq_softmax

* use cub in seq_softmax

* Improve and fix fake_quantize_op.

* cudnn widndows

* "add comment"

* "windows support"

* "fix cmake error"

* add flatten op

* refine regularization for selected_rows

* clean lookup_table

* refine rpc_server_test

* temporally disable rpc_server_test

* fix rpc_server_test

* add unit test

* also convert tab to space

* "split into multiple .ccl"

* "refine file structure"

* "refine files"

* "remove the cmakelist"

* "fix typo"

* "fix typo"

* fix ci

* Develop a fake dequantized op for fixed-point quantization training framework.

* Add the missing file.