Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

关于重参数化reparam.py文件 #21

Open
unihuihui opened this issue Apr 23, 2024 · 4 comments
Open

关于重参数化reparam.py文件 #21

unihuihui opened this issue Apr 23, 2024 · 4 comments

Comments

@unihuihui
Copy link

您好,我有几个疑问烦请解答一下。
1、在您提供的reparam.py文件中line40的变量saved_model_path指代的权重文件和您放在百度网盘中权重文件不一样,请问有什么区别吗?
2、line42的ckp=ckp['model'],但是在debug过程中没看见权重文件中有这个变量?另外这句代码表达什么什么含义?
3、我查看了您百度网盘中提供的三个权重文件,都没有conv1_1、conv1_2、conv1_3、conv1_4、conv1_5?
4、还是说,您百度网盘中提供的权重文件已经是经过重参数化后的权重文件

@unihuihui
Copy link
Author

@cecret3350

@cecret3350
Copy link
Owner

您好,

  1. line40的变量saved_model_path指代的权重文件为未重参数化的权重,而百度网盘中的权重文件已经经过重参数化
  2. reparam.py针对的是利用train.py训练的模型,请查看train.py中保存模型的代码
  3. 因为百度网盘中的权重文件已经经过重参数化,所以conv1_1、conv1_2、conv1_3、conv1_4、conv1_5已经被合并为一个等效的常规卷积
  4. 是的

@unihuihui
Copy link
Author

谢谢您

@tianyalei
Copy link

image
作者您好,我看您的代码中的细节卷积采样上述代码,您在训练和推理的时候都是这个代码吗?那重参数是什么意思呢,跟这个有啥关系吗

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants