- It provides a unified style to express graph and kernel for user.
- It provides a unified IR to represent graph and kernel for developer.
- It breaks the boundary between graph and kernel.
- It provides more opportunities to do compile optimization.

This reverts commit 705c71a257.

Merge pull request !1785 from lianliguang/add-dropout-kernel-special-kernel-select-rules

Merge pull request !1794 from dengwentao/cpu_op

* add infer_value for common math ops
* convert constant bool tensor to bool value

* do not infer value when encounter 0 as division

for while condition, do not unrool if condition is a tensor

Merge pull request !1578 from gengdongjie/master

This reverts commit 3db8cfa54f.

Merge pull request !1160 from wangqiuliang/add-backward-hook-in-pynative-mode

Signed-off-by: zhoufeng <zhoufeng54@huawei.com>

fix DiagPart constant folding issue

fix argmin output type check

fix atan2 doc error

fix remove FusedBatchNorm and its grad

This reverts commit ea6958c50a.

Merge pull request !340 from ougongchang/master

Merge pull request !246 from vlne-v1/I1E3PI-opt-pass-adjust-allreduce-apply-weight-decy-seq

clean clang format errors and cpplint errors

add some test cases for histogram summary operator

Signed-off-by: leonwanghui <leon.wanghui@huawei.com>