* split yolov3 loss using paddle op

* add sync_bn and modify list to LayerList to use sync_bn, add missing op, support cutmix op in dataset, modify reader to use new ops

* modify code according to review

* modify code to run eval.py successfully

* modify code to run eval and train successfully

* modify code to use mixup

* rebase code on lastest dygraph

* modify code to run in sync_bn mode

* modify code according to review

* modify target size of ResizeOp

* update mask_rcnn_fpn model

* update smooth_l1_loss

* update smooth_l1 and sigmoid_cross_entropy_with_logits

* fix relu name

* update reader in dygraph

* update dataloader some detail

* fix num_classes with background

* rename GridMaskOp to GridMask in ppdet/data/transform/operator.py to avoid conflict

* fix bugs in operator and batch_operator and check_version

* modify name in GridMask

* add comma according to review

* split config into models, optimizers, dataset, readers，runtime

* modify operators and batch_operators in data preprocess

* import BboxError and ImageError in operators

* modify code according to review

* modify some bugs

* add Gt2Solov2TargetOp in batch_operators

* modify code according to review

* update YOLOv3 config

* add matrix nms api

* add test error in matrix nms ut

* modify code according to review

* modify matrix nms api

* add yolo_box and add unittest of yolo_box.

* update comment

* update sample code, call op in ppdet

* migrate roi_align and roi_pool to ops

* modify code according to review

* modify code to be suited for launch in 2.0rc

* modify code according to review

* migrate detection ops from Paddle to PaddleDetection

* add unittest for detection ops

* update api & unittest

* add copyright & api name to migrate

* refine PaddleDetection benchmard print, test=develop

* refine PaddleDetection benchmard print, test=develop

* refine PaddleDetection benchmard print, test=develop

* refine PaddleDetection benchmard print, test=develop

* refactor dygraph & add mask rcnn fpn

* fix initialization & minor update

* refine architecture of mask_rcnn

split target op into label&sample op