Merge pull request !4629 from lichen_101010/sendgraph_sendtensor_issue

Merge pull request !4860 from 徐永飞/master

Merge pull request !4882 from zhaodezan/master

Merge pull request !4880 from VectorSL/model-zoo

Merge pull request !4881 from byweng/fix_param_check

Merge pull request !4864 from zhanyuan/dev

Merge pull request !4691 from linqingke/new_ops

Merge pull request !4789 from liuxiao93/Add-EditDistance-op-for-GE

Merge pull request !4801 from yihuaijie/dev

Merge pull request !4863 from luoyang/c-api-pyfunc

Merge pull request !4743 from yuximiao/master

Merge pull request !4846 from yuximiao/gpu_comments

Merge pull request !4820 from shenwei41/sw_master

Merge pull request !4830 from yuchaojie/gpu_allreduce

Merge pull request !4845 from zhangxinfeng3/master

Merge pull request !4838 from wangmin0104/master

Merge pull request !4874 from peixu_ren/custom_bijector

Merge pull request !4875 from XunDeng/pp_issue_branch

!4871 GE backend does not use glog (-g off) and is causing linking error on linking cache_server exe
Merge pull request !4871 from JesseKLee/tdt

Merge pull request !4472 from anthonyaje/gpu_queue_size

change buffer type and chunk size

addressed sendtensor issue for large tensors

fix sendtensor part 2

fix bugs for sendtensor issue

fix CI errors

removes a print line

address comments

Merge pull request !4706 from Peilin/smoothL1Loss-fix

Merge pull request !4314 from nhussain/clean_up_bindings

Merge pull request !4683 from yoonlee666/cisplit

Merge pull request !4771 from EricZ/md_tensor_from_mem

Merge pull request !4843 from xychow/remove-ccsrc-common-h

Merge pull request !4475 from JonathanY/batchnormgrad

Merge pull request !4857 from tom_chen/deepfm_lr

Merge pull request !4858 from zhaodezan/master

Merge pull request !4819 from zyli2020/Add_FusedBatchEx_support