yjh0410 88eab2141c modify the link to pretrained weight		2 жил өмнө
..
README.md	88eab2141c modify the link to pretrained weight	2 жил өмнө
build.py	3f8c6ecea7 add ONNX deployment	2 жил өмнө
loss.py	0744feec35 add YOLOv7-Plus	2 жил өмнө
matcher.py	7cf531e7da add Tracking	2 жил өмнө
yolov3.py	3f8c6ecea7 add ONNX deployment	2 жил өмнө
yolov3_backbone.py	b9f295539e keep training YOLOvx-S	2 жил өмнө
yolov3_basic.py	7cf531e7da add Tracking	2 жил өмнө
yolov3_fpn.py	7127f065b3 add YOLOv4-Tiny	2 жил өмнө
yolov3_head.py	7cf531e7da add Tracking	2 жил өмнө
yolov3_neck.py	7cf531e7da add Tracking	2 жил өмнө

YOLOv3:

For training, we train YOLOv3 and YOLOv3-Tiny with 250 epochs on COCO.
For data augmentation, we use the large scale jitter (LSJ), Mosaic augmentation and Mixup augmentation, following the setting of YOLOv5.
For optimizer, we use SGD with momentum 0.937, weight decay 0.0005 and base lr 0.01.
For learning rate scheduler, we use linear decay scheduler.
For YOLOv3's structure, we use decoupled head, following the setting of YOLOX.

Model	Backbone	Batch	Scale	AP^val 0.5:0.95	AP^val 0.5	FLOPs ^(G)	Params ^(M)	Weight
YOLOv3-Tiny	DarkNet-Tiny	1xb16	640	25.4	43.4	7.0	2.3	ckpt
YOLOv3	DarkNet-53	1xb16	640	42.9	63.5	167.4	54.9	ckpt