Stinky-Tofu / Stronger-yolo

🔥Improve yolo with latest paper
MIT License
3 stars 0 forks source link

Dose cosine decay speed up the training? #55

Open gooodooo opened 5 years ago

gooodooo commented 5 years ago

I noticed that the original YOLOv3 uses steps decay,and your v2 has a faster training than the original YOLOv3. Or just because you use tensorflow and the original use darknet?

Stinky-Tofu commented 5 years ago

@gooodooo I don't know ...

chestnutLeeeee commented 5 years ago

I think Adam works,original yolov3 use SGD