Merge pull request !3620 from 张学同/to_merge

Merge pull request !3270 from thlinh/dev_Jul17_removeSelect

Merge pull request !3612 from changzherui/mod_case

Refactored MapOp to in preparation to support device augment

Minor fix

Merge pull request !3394 from XunDeng/pp_poc_v3

Merge pull request !3352 from JonathanY/ROIAlignGrad

Merge pull request !3560 from peixu_ren/custom_gpu

Merge pull request !3551 from nhussain/fix_cpp_only_build

Merge pull request !3543 from lianliguang/unify-primitive

Merge pull request !3571 from guansongsong/gss/fix_cache_core

Merge pull request !3453 from zichun_ye/resolve_attr_pr

Create a new pass named ValueBasedEliminate to reduce the load of Arithmetic Simplify

Code review

update logic for cache

clean up code

add test cases for new api

update include of code

specify the resolve attr to ClassMember

avoid resolve attr of self

Merge pull request !3472 from Xiaoda/13-add-gpu-costmodel

Merge pull request !3528 from Simson/push-to-opensource

Merge pull request !3435 from yangzhenzhang/add-tile-op

added cdf, logcdf, survival, logsurvival, entropy, mode, cross_entropy wrapper

Merge pull request !3449 from MahdiRahmaniHanzaki/cifar100-c-api

merge the script

optimize the script

Merge pull request !3519 from anzhengqi/remove-log

Merge pull request !3500 from lirongzhen1/master

Merge pull request !3525 from hewei/fix_parameter_bug

Merge pull request !3511 from linqingke/new_ops

1. Parameter's init_data() should have no effect if default_input already set;
2. This bug is introduced by 'decouple ParamValue from python';
3. An unit test case is added to ensure the right init_data() behavior.

Merge pull request !3513 from ZPaC/master-add-ps-test-cases

Merge pull request !3425 from fangzehua/avgpool

Merge pull request !3497 from yuchaojie/transformer_st

Merge pull request !3349 from liubuyu/master

Merge pull request !3495 from chenweifeng/strided_slice

Merge pull request !3324 from fangzehua/reduce_any