yjh0410 a7d44da089 update README		2 anni fa
..
README.md	a7d44da089 update README	2 anni fa
build.py	73011ee555 debug YOLOX-style Transform with Rotation	2 anni fa
loss.py	efa0bed448 debug YOLOX-style Transform with Rotation	2 anni fa
matcher.py	2727b6adb1 debug trainer	2 anni fa
yolox.py	73011ee555 debug YOLOX-style Transform with Rotation	2 anni fa
yolox_backbone.py	8f4509bd61 modify YOLOX	2 anni fa
yolox_basic.py	3246f3efdd update	2 anni fa
yolox_head.py	311a9b89b7 fix a unknown bug in YOLOX	2 anni fa
yolox_neck.py	8f4509bd61 modify YOLOX	2 anni fa
yolox_pafpn.py	3246f3efdd update	2 anni fa

YOLOv5:

For training, we train YOLOX series with 300 epochs on COCO.
For data augmentation, we use the large scale jitter (LSJ), Mosaic augmentation and Mixup augmentation, following the setting of YOLOX.
For optimizer, we use AdamW with weight decay 0.05 and base per image lr 0.001 / 64.
For learning rate scheduler, we use linear decay scheduler.
I am trying to retrain YOLOX-M and YOLOX-L with more GPUs, and I will update the AP of YOLOX-M and YOLOX-L in the table in the future.

Model	Backbone	Batch	Scale	AP^val 0.5:0.95	AP^val 0.5	FLOPs ^(G)	Params ^(M)	Weight
YOLOX-N	CSPDarkNet-N	8xb8	640	30.4	48.9	7.5	2.3	ckpt
YOLOX-S	CSPDarkNet-S	8xb8	640	39.0	58.8	26.8	8.9	ckpt
YOLOX-M	CSPDarkNet-M	1xb16	640	44.6	63.8	74.3	25.4	ckpt
YOLOX-L	CSPDarkNet-L	1xb16	640	46.9	65.9	155.4	54.2	ckpt