Merge pull request !2513 from laiyongqiang/r0.5

Merge pull request !2525 from liuwenhao/r0.5

Merge pull request !2523 from zhoufeng/disable-cuda9-r0.5

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

Merge pull request !2509 from zhoufeng/graph-kernel-use-control-sink-r0.5

Merge pull request !2493 from YuJianfeng/r0.5

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

Merge pull request !2476 from ougongchang/fix_folding_constant_r0.5

Merge pull request !2488 from mindspore_ding/backend

Merge pull request !2467 from chenfei_mindspore/check-depend-mode-of-control-depend

Merge pull request !2481 from dinghao/r0.5

Merge pull request !2461 from YuJianfeng/r0.5

The summary operator will be optimized when it return the origin value
in constant folding scene. So I return a None value to avoid this.

Merge pull request !2462 from kisnwang/optimize-reduce-sparse-gradient

Merge pull request !2449 from chenfei_mindspore/fix-bug-ascend-control-parser

Merge pull request !2456 from zhoufeng/xiu-ba-ge

Merge pull request !2453 from ougongchang/r0.5

Merge pull request !2444 from kisnwang/fix-mix-target-entry-for-r0.5

Merge pull request !2433 from qianlong21st/add_text_comment

Merge pull request !2413 from Xiaoda/7-auto-parallel-check-optimizer-ops

Merge pull request !2412 from jiangjinsheng/issue_fix4

Merge pull request !2427 from TronZhang/akg_default_on

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

Merge pull request !2417 from liuwenhao/master

Merge pull request !2177 from jiangjinsheng/vm_lin_space

Merge pull request !2424 from xutianchun/ctcloss

Merge pull request !2369 from baihuawei/reduce

Merge pull request !2398 from h.farahat/slice_bug

Merge pull request !2404 from xulei/input_output

Merge pull request !2375 from h.farahat/python_tokenizer

Merge pull request !2420 from liyong126/mindrecord_epoch_shuffle

Merge pull request !2393 from lianliguang/fix-bug-of-merge-cast-to-op-and-gene-hccl-kernel-info