2- Search about the Tensorboard and Evaluations afterward:

[m-kashani]

• https://github.com/facebookresearch/detectron2/blob/master/detectron2/structures/boxes.py#L299

---

├─evaluation <- evaluator for each dataset

https://github.com/facebookresearch/detectron2/tree/master/detectron2/evaluation

[ ] - To do first (3-4h)

├─engine <- predictor and trainer engines

https://github.com/facebookresearch/detectron2/tree/master/detectron2/engine

├─model_zoo <- pre-trained model links and handler
https://github.com/facebookresearch/detectron2/tree/master/detectron2/model_zoo

Faster RCNN

Backbone	Param.	lr sched	inference time	AP	APs	APm	APl	download
V2-19-FPN
R-50-FPN
V2-39-FPN
4
5

Mask R-CNN

Backbone	lr sched	inference time	box AP	box APs	box APm	box APl	mask AP	mask APs	mask APm	mask APl	download
V2-19-FPNLite
V2-19-FPN
V2-39-FPN
3

[m-kashani]

Panoptic-FPN on COCO

Name	lr sched	inference time (s/im)	box AP	mask AP	PQ	download
R-50-FPN	..	..	..	..	..	model | metrics
V2-39-FPN	..	..	..	..	..	..
R-101-FPN	..	..	..	..	..	..
V2-57-FPN	..	..	..	..	..	..