Merge pull request !2720 from xychow/fix-assign-in-while

- 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.

Merge pull request !2024 from Kang/master

support tensor slice using constexpr

remove tensorslice metagraph

add pynative testcases

Merge pull request !1860 from fary86/fix_codex_06.05

* 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 !1803 from Kang/master

Merge pull request !1766 from vlne-v1/I1J0M0-amp-do-auto-cast-failed

Merge pull request !1385 from amongo/SupportMultiSwitch

Merge pull request !1433 from leopz/tensor_base

resolve review

modify infer to get abstract value and attr value map together

support get attr in specialize

clone primitive when specialize if attribute is not the same with
infered attribute

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

add some unittest for summary operator

support ScalarSummary to record number type

Merge pull request !1142 from ZhangQinghua/master

Merge pull request !1120 from Kang/master

Merge pull request !683 from xychow/bypass-renorm-and-specialize-hypermap-parameter

Merge pull request !721 from candanzg/tensor_assgin_ellipsis

Merge pull request !814 from xychow/compare-with-context-ptr

Signed-off-by: candanzg <zhangshucheng@huawei.com>

Merge pull request !700 from penn/validate_bprop_rules

Merge pull request !731 from fary86/remove_extra_empty_string_in_log

Merge pull request !723 from Simson/push-to-opensource