Syencil mobile-yolov5-pruning-distillation issues

Syencil / mobile-yolov5-pruning-distillation

mobilev2-yolov5s剪枝、蒸馏，支持ncnn，tensorRT部署。ultra-light but better performence！

MIT License

816 stars 164 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

AttributeError: 'Model' object has no attribute 'names'

#82 YongpeiZhu opened 1 year ago
0
'python3 train.py --type vocl' meet the error

#81 ResidualNS opened 1 year ago
0
蒸馏代码问题

#80 ftl123456789 opened 1 year ago
0
蒸馏模块无法跑起来

#79 nayenizhenmei opened 1 year ago
2
关于train后识别不到任何结果

#78 yakexi-cmd closed 1 year ago
0
model文件夹中yaml文件

#77 zy597337447 opened 1 year ago
0
Update LICENSE

#76 victor30608 opened 1 year ago
0
feature蒸馏的一些疑惑

#75 xxxwuwq closed 2 years ago
0
no fuse

#74 gy-zkj opened 2 years ago
0
求助帖

#73 DoAlone opened 2 years ago
3
旋转框蒸馏

#72 xing-bing closed 2 years ago
6
蒸馏后Student Model的准确率下降

#71 WenzhiShao opened 2 years ago
1
请问如何配置apex进行混合多精度加速训练

#70 mengxiangqiqin opened 2 years ago
0
作者你好，请问剪枝过程是参考的哪篇论文

#69 leiqing110 closed 2 years ago
0
大佬好，ignore_idx 是不是写错了

#68 Veal98 closed 2 years ago
1
before_pruning.jpg和after_pruning.jpg

#67 yuzhixiaobo closed 2 years ago
0
关于TA策略与d_online

#66 StrongerzZ opened 2 years ago
1
大佬好，注意到 yolo head 和原版的有所不同，想问下具体做了哪些改变

#65 Veal98 closed 2 years ago
4
AttributeError: Can't get attribute 'ModelPruned' on <module 'models.yolo' from 'models\\yolo.py'>

#64 hzc5095 opened 2 years ago
0
模型剪枝之后，参数减少很多，但是模型推理速度并没有明显区别

#63 Ylifei opened 2 years ago
0
稀疏训练时遇到一些问题

#62 yezechen opened 2 years ago
1
大佬好，求教小白如何顺利运行，感觉缺失了一些数据文件

#61 Veal98 closed 2 years ago
1
pytorch版本

#60 Xingxu1996 opened 3 years ago
0
mAP值一直不变，从一开始就是接近0

#59 wxz1996 opened 3 years ago
4
关于mobile-yolov5s的mAP

#58 EEEricLiu opened 3 years ago
2
权重问题

#57 1012207826 opened 3 years ago
0
mobilenet的预训练权重

#56 jzh-Joshua closed 3 years ago
1
How to fix this problem 'Detect'object has no attribute 'm'

#55 Niki173 opened 3 years ago
0
Mobilev2-Yolov5 的throughput

#54 EEEricLiu opened 3 years ago
1
请问bn参数分布图如何绘制

#53 qq302639209 opened 3 years ago
0
权重文件不能下载

#52 MoXQian opened 3 years ago
3
使用python3 train.py --type mvocs训练出来的pt模型是20.7M，这正常吗？我看baseline的参数量是3.60 Params(M)，这是为什么呢

#51 tommy3266 opened 3 years ago
2
、

#50 chengzi999 closed 3 years ago
0
1

#49 zack2020-star closed 3 years ago
0
请问稀疏参数的选取有什么原则吗？

#48 15050188022 opened 3 years ago
0
可以私聊点问题吗

#47 neasfss opened 3 years ago
1
这个可以扩展至5x吧？

#46 Bobby-youngking opened 3 years ago
0
关于teacher model预训练权重

#45 Sharpiless opened 3 years ago
1
遇到了点问题

#44 Lihong1997 opened 3 years ago
0
蒸馏

#43 yyyooohao opened 3 years ago
4
'Got {}, but numpy array, torch tensor, or caffe2 blob name are expected.'.format(type(x))) NotImplementedError: Got <class 'NoneType'>, but numpy array, torch tensor, or caffe2 blob name are expected.

#42 ahhhxxx opened 3 years ago
2
RuntimeError: shape '[1, 3, 85, 80, 64]' is invalid for input of size 655360

#41 ahhhxxx opened 3 years ago
3
关于finetune

#40 15050188022 opened 3 years ago
2
策略

#39 yyyooohao closed 3 years ago
0
常见的一些问题汇总

#38 Syencil opened 3 years ago
3
关于剪枝蒸馏的策略

#37 15050188022 closed 3 years ago
2
请问一下，使用预训练的yolov5s在coco上做基础训练，验证集的map非常接近于0，问题大概出在什么地方？

#36 August0424 closed 3 years ago
2
剪枝之后的模型也可以通过onnx转换和通过ncnn部署吗？

#35 zineos closed 3 years ago
1
蒸馏时Precision下降太多的问题

#34 CangHaiQingYue closed 3 years ago
2
L1 稀疏化训练细节

#33 xuanyuyt closed 3 years ago
0