Eric-mingjie / network-slimming

Network Slimming (Pytorch) (ICCV 2017)
MIT License
909 stars 214 forks source link

如何进行迭代剪枝 #28

Open liulonghoi opened 5 years ago

liulonghoi commented 5 years ago

文中所提到的迭代剪枝是将您提供的代码进行一次,然后将得到微调后的模型重复train with sparsity ,prune等过程吗?我实际去做的时候发现剪枝一次后的模型再重复剪枝的话减掉的通道和第一次剪枝一样,并没有剪枝掉更多的通道,这是什么原因呢?

Eric-mingjie commented 5 years ago

The current code is designed to support one-shot pruning. Please take a look at the pruning code. I think for VGG you can reuse the pruning code for iterative pruning. For ResNet and DenseNet, there is a channel selection layer. Therefore, you can not reuse the pruning code for iterative pruning.

liulonghoi commented 5 years ago

Thank you for your reply,could you show me the puring code?

Eric-mingjie commented 5 years ago

We don't have it yet.

HouBiaoLiu commented 4 years ago

我按照作者的改了一个迭代剪枝的,可进行再次剪枝

tangbohu commented 4 years ago

我按照作者的改了一个迭代剪枝的,可进行再次剪枝

可否分享下code,email address: tangbohutbh@gmail.com

17868380981 commented 4 years ago

我按照作者的改了一个迭代剪枝的,可进行再次剪枝

能分享一下迭代剪枝的代码嘛。

17868380981 commented 4 years ago

我按照作者的改了一个迭代剪枝的,可进行再次剪枝

1109720378@qq.comqq.com

zmzds commented 2 years ago

我按照作者的改了一个迭代剪枝的,可进行再次剪枝

能分享一下代码吗?810937616@qq.com

17868380981 commented 2 years ago

不好意思,我已经毕业了,代码要丢了,你看有没有通讯作者可以联系一下。

---原始邮件--- 发件人: @.> 发送时间: 2022年8月9日 21:15:48 收件人: @.>; 抄送: @.**@.>; 主题: Re: [Eric-mingjie/network-slimming] 如何进行迭代剪枝 (#28)

我按照作者的改了一个迭代剪枝的,可进行再次剪枝

@.***

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you commented.Message ID: @.***>

zmzds commented 2 years ago

不好意思,我已经毕业了,代码要丢了,你看有没有通讯作者可以联系一下。

那可以分享一下可参考的代码吗?通讯作者好像联系不到