iamhankai ghostnet.pytorch issues

iamhankai / ghostnet.pytorch

[CVPR2020] GhostNet: More Features from Cheap Operations

https://arxiv.org/abs/1911.11907

522 stars 118 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

关于卷积核尺寸的疑问

#40 Tinkle0 opened 4 years ago
1
我将Ghostnet用在resnet中，还需要对网络进行训练吗？

#39 Tangzhaotz closed 4 years ago
2
用GhostModule替换残差50的Bottleneck中的nn.Conv2d报错

#38 Tangzhaotz closed 4 years ago
2
ghostnet用于细粒度图像分类

#37 Javacr opened 4 years ago
3
test speed flops and parameters

#36 Amarintine opened 4 years ago
2
用ghostnet做anchorfree目标检测的主干网络

#35 gp12345678 opened 4 years ago
5
How to set the channels?

#34 DYJ1111 closed 4 years ago
4
为什么我在替换nn.Conv2d的时候，初始化权重会报错？

#33 tianyuluan opened 4 years ago
5
The neural network slowed down with only replace the Conv2d in my net with GhostModule.

#32 JacobLau84 closed 4 years ago
11
Params/ FLOPs counter

#31 digantamisra98 closed 4 years ago
1
Can not run the sample

#30 DYJ1111 closed 4 years ago
6
kernet is kernel

#29 freaad closed 4 years ago
1
Robustness to linear quantization

#28 wzds2015 closed 4 years ago
1
About the accuracy of Ghost-ResNet-50 (s=2) on Imagenet

#27 JiaxiongQ closed 4 years ago
1
kernet_size in ghost_net.py line 101, it should be kernel_size

#26 aksenventwo closed 4 years ago
2
Stride=1 cfg Kernel Sizes Never Used?

#25 glenn-jocher closed 4 years ago
2
Custom Weight Initialization

#24 glenn-jocher opened 4 years ago
3
Update ghost_net.py

#23 glenn-jocher closed 4 years ago
0
Replace nested nn.Sequential() with nn.Identity()

#22 glenn-jocher closed 4 years ago
2
the activation fn of depthwiseconv in res connectionis different from that of tensorflow version??

#21 hurri2000 closed 4 years ago
6
请问，可否使用您提供的预训练模型进行finuting，可否提供下您训练时使用的预处理

#20 duducode closed 4 years ago
0
cheap_operation

#19 Abeni18 closed 4 years ago
2
Loss and Epocs

#18 dronefreak closed 4 years ago
4
Release of PyTorch pretrained model and ablation

#17 d-li14 closed 4 years ago
3
How to init GhostModule

#16 KopiSoftware closed 4 years ago
2
hello! how to move the GhostModule to my net to replace the normal conv2d layer?

#15 tianyuluan opened 4 years ago
4
Can the activation function be changed to prelu?

#14 ZhangYang0726 closed 4 years ago
1
关于论文中ResNet-56和Ghost-ResNet-56训练参数问题

#13 breeze0 closed 4 years ago
4
关于cheap conv中relu的使用问题

#12 pawopawo closed 4 years ago
9
Why are the k*k depthwise conv only used for downsample?

#11 LightToYang closed 4 years ago
1
cfgs for cifar-10

#10 laisimiao closed 4 years ago
1
cheap_opt is liner in paper, why relu?

#9 zhaowenlovehome closed 4 years ago
4
How to train GhostNet? You didn't mention. Plz..

#8 ArtificialZeng closed 4 years ago
1
Ghost-VGG-16和Ghost-ResNet-56的训练参数

#7 breeze0 closed 4 years ago
6
width_multi

#6 pawopawo closed 4 years ago
1
pretrained pth model for pytorch

#5 dragen1860 closed 4 years ago
4
如何应用到resnet模型？

#4 splinter21 closed 4 years ago
1
pytorch训练的超参数是？

#3 betterhalfwzm closed 4 years ago
12
About SELayer

#2 WangDeyu closed 4 years ago
4
small version will be released soon? Did you do some experiments on it ?

#1 DuckJ closed 4 years ago
8