YOLOv3-complete-pruning
YOLOv3-complete-pruning copied to clipboard
coldlarry
剪枝后map=0
作者你好，稀疏训练后，模型大小为420M,然后剪枝,MAP=0,这是咋回事啊 +------------+----------+----------+ | Metric | Before | After | +------------+----------+----------+ | mAP | 0.797547 | 0.000000 | | Parameters | 61523734 | 9693894 | | Inference | 0.1313 | 0.0852 | +------------+----------+----------+ Config file has been saved: cfg/prune_0.85_yolov3-hand.cfg Compact model has been saved: weights/yolov3_hand_regular_pruning_0.85percent.weights
剪枝率设置的太高
------------------ 原始邮件 ------------------ 发件人: "yangdeqieyingli"<[email protected]>; 发送时间: 2019年11月14日(星期四) 下午3:31 收件人: "coldlarry/YOLOv3-complete-pruning"<[email protected]>; 抄送: "Subscribed"<[email protected]>; 主题: [coldlarry/YOLOv3-complete-pruning] 剪枝后map=0 (#21)
作者你好，稀疏训练后，模型大小为420M,然后剪枝,MAP=0,这是咋回事啊 +------------+----------+----------+ | Metric | Before | After | +------------+----------+----------+ | mAP | 0.797547 | 0.000000 | | Parameters | 61523734 | 9693894 | | Inference | 0.1313 | 0.0852 | +------------+----------+----------+ Config file has been saved: cfg/prune_0.85_yolov3-hand.cfg Compact model has been saved: weights/yolov3_hand_regular_pruning_0.85percent.weights
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or unsubscribe.
恩恩，已经好啦，谢谢你，但是稀疏训练后模型大小400多M正常吗？ 只有当手出现在特定距离时才会比较准确的识别，太近，太远，都识别不到？
------------------ 原始邮件 ------------------ 发件人: "yangdeqieyingli"<[email protected]>; 发送时间: 2019年11月14日(星期四) 下午3:31 收件人: "coldlarry/YOLOv3-complete-pruning"<[email protected]>; 抄送: "Subscribed"<[email protected]>; 主题: [coldlarry/YOLOv3-complete-pruning] 剪枝后map=0 (#21)
稀疏后420M？还是420MB？ 两个好像都不对啊
作者你好，稀疏训练后，模型大小为420M,然后剪枝,MAP=0,这是咋回事啊 +------------+----------+----------+ | Metric | Before | After | +------------+----------+----------+ | mAP | 0.797547 | 0.000000 | | Parameters | 61523734 | 9693894 | | Inference | 0.1313 | 0.0852 | +------------+----------+----------+ Config file has been saved: cfg/prune_0.85_yolov3-hand.cfg Compact model has been saved: weights/yolov3_hand_regular_pruning_0.85percent.weights
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or unsubscribe.
python3 train.py --data data/oxfordhand.data --batch-size 32 --accumulate 1 --weights weights/yolov3.weights --cfg cfg/yolov3-hand.cfg -sr --s 0.001 --prune 0 按照这个训练得到的模型是420M，正常吗？
python3 train.py --data data/oxfordhand.data --batch-size 32 --accumulate 1 --weights weights/yolov3.weights --cfg cfg/yolov3-hand.cfg -sr --s 0.001 --prune 0 按照这个训练得到的模型是420M，正常吗？
你好 我也遇到了同样的问题 这个是因为保存了optimizer 的梯度数据 可以运行utils.py中
