* fix yolov3 voc eval in train

* fix comment

* standard SENet and add a warning in coco loader

* add deformable models
* fix dcn v2
* fix log_iter

* fix easy problem

Add PaddleDetection docs & some model configs

* add voc_eval and yolo_darknet_voc

* add yolov3_darknet_voc in MODEL_ZOO

* fix default im_size

* fix MODEL_ZOO note

* fix is_bbox_normalized

* extract map to map_utils.py

* update yolov3_dorknet_voc mixup

* add yolov3_r34_voc

* add yolov3_mobilenet_v1_voc

* fix drop empty in VAL mode

* use cfg.num_classes

* assert metric valid

* enable difficulty can be None

* add comment for bbox_eval

* num_classes in retinanet

* Fix ResizeImage and data transform for SSD
* Fix code style

option should not be overridden by subsequent options of the same module.

* fix infer doc and config default

* refine dataset exists

* fix None proc

* add rpn recall function and class-aware rpn

* optimize error message when result is none

* fix resnet for resnet34

* fix ResNet18 and ResNet34

* update modelzoo and fix configs

* fix configs

* fix resnet bn

* Clean up a bit

* Add `__shared__` annotation for shared variables between modules

* Dedup `num_classes` configurations

* Document `__shared__` usage

* Remove some unused variable

* Improve docstring and add comments for `__shared__`

some of the global config values are reference type, e.g., objects, lists or
dicts, if the created module instances modify those values (see FPN and
spatial_scale), global config will reflect these changes, and instances of the
same class created later will inherit the changed values

* refine bbox_normalize in infer.py

* add is_bbox_normalize

* rename _forward to build

* check callable

* check gpu for ppdet/yolov3

* add random shape descp in MODEL_ZOO

* refine doc

* not use core

* Rename object_detection to PaddleDetection
* Small fix for doc