Problem: Create 'uint8' type tensor from float data raise RuntimeError;
Fix:  Use same internal data type for 'uint8' and 'bool' tensor and allows
convert float data to bool tensor.

Merge pull request !2787 from zhaozhenlong/op/scatter-div-mul-vm

Merge pull request !2794 from zhouyuanshen/master

Merge pull request !2700 from chujinjin/change_pynative_batchnorm_same_as_graph_mode

Merge pull request !2750 from TronZhang/clean_bot

Merge pull request !2780 from Margaret_wangrui/handle_partial

Merge pull request !2792 from yanghaitao/yht_fastrcnn_accu

Merge pull request !2763 from yuchaojie/transformer

Merge pull request !2777 from chenweifeng/lstm_performace

Merge pull request !1934 from ZPaC/add-pslite-dependency

Merge pull request !2788 from yangzhenzhang/delete-shape-validation

Merge pull request !2783 from dinghao/master

Merge pull request !2782 from zhoufeng/control-sink-test

Merge pull request !2786 from Kang/optimization

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

!2771 Address the importance of properly closing the SummaryRecord and illustrate how to close in two ways
Merge pull request !2771 from LiHongzhang/improve_docs

Merge pull request !2638 from jiangjinsheng/vm_mod

Merge pull request !2767 from wangqiuliang/fix-hook-operator-compare-issue

Merge pull request !2722 from Kang/optimization

Merge pull request !2747 from zhangbuxue/refactor_the_StridedSlice_op

Merge pull request !2514 from yoonlee666/gelu2

Merge pull request !2769 from ougongchang/fix_nonetype

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

Merge pull request !2775 from ZiruiWu/better_get_dataset_size_in_generator

Merge pull request !2772 from Jamie/removalpass

cpplint