yolov5_prune
yolov5_prune copied to clipboard
train with channel sparsity regularization
这个参数是必须的吗?可否只需要指定--scale --prune
without the parameters,how to define a train action?