prune后mAP直接为0 - Githubissues

SpursLipu / YOLOv3v4-ModelCompression-MultidatasetTraining-Multibackbone

YOLO ModelCompression MultidatasetTraining

GNU General Public License v3.0

444 stars 136 forks source link

prune后mAP直接为0 #118

Closed Biubiubiu12 closed 3 years ago

Biubiubiu12 commented 3 years ago

python3 normal_prune.py --cfg cfg/yolov3/yolov3-hand.cfg --data data/oxfordhand.data --weights weights/best.pt --percent 0.5 +------------+----------+----------+ | Metric | Before | After | +------------+----------+----------+ | mAP | 0.764819 | 0.000000 | | Parameters | 61523734 | 21907584 | | Inference | 0.0902 | 0.0658 | +------------+----------+----------+ second step --s 0.001,将稀疏后的best.pt进行裁剪后，mAP为0 我将percent缩小至0.1,仍然不起作用

MohamedElsaeidy commented 3 years ago

@Biubiubiu12 i have same issue could you solve ?

Biubiubiu12 commented 3 years ago

must spare train before pruning

---原始邮件--- 发件人: @.> 发送时间: 2021年4月27日(周二) 晚上8:37 收件人: @.>; 抄送: @.**@.>; 主题: Re: [SpursLipu/YOLOv3v4-ModelCompression-MultidatasetTraining-Multibackbone] prune后mAP直接为0 (#118)

@Biubiubiu12 i have same issue could you solve ?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.