PyTorch_YOLOv4
PyTorch_YOLOv4 copied to clipboard
您好,这个支持darknet版的yolov4-p5和p6的cfg吗?
https://github.com/AlexeyAB/darknet/blob/master/cfg/yolov4-p6.cfg
yes, just change the activation before yolo layers from logistic to linear and remove some unsupported fields in yolo layers.
Hi @WongKinYiu , are all the provided weights in this repo are with the logistical wh in yolo layers? so the yolov4.weights is different from AlexeyAB's one? Or only those pacsp models are with the logistical wh?
@atangfan The .weights-files are identical in https://github.com/AlexeyAB/darknet and https://github.com/WongKinYiu/PyTorch_YOLOv4 but different cfg-files:
- for Darknet you should use
activation=logistic
- for PyTorch_YOLOv4 you should use
activation=linear
Where can we get the pt files? @AlexeyAB @WongKinYiu Since train.py requires only those.