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

代码训练问题,权重最优,图像融合 #12

Open
guigui6666 opened this issue Aug 6, 2023 · 4 comments
Open

代码训练问题,权重最优,图像融合 #12

guigui6666 opened this issue Aug 6, 2023 · 4 comments

Comments

@guigui6666
Copy link

您好!
我看了您的代码,想请问您下面两个问题:

  1. 请问您是如何保存效果最优的那一轮权重的?代码里好像只保存了最后一轮的权重,但似乎有时候前面的权重会比后面的权重测试效果更好。
  2. 请问为何您的图像融合部分不放在训练代码中,直接端到端的输出,而要放在测试代码里,有什么理由么?
    期待您的回复。
@xiaofeng94
Copy link
Owner

xiaofeng94 commented Aug 6, 2023

Hey, 非常感谢关注,以下是我们的回答:

  1. 我们并没有保存最优权重。并且不建议根据测试结果挑选最优权重,那可能会导致数据泄露和过拟合
  2. 我们的图像融合方法是一个离线过程,不需要和特定的训练过程绑定

@guigui6666
Copy link
Author

guigui6666 commented Aug 7, 2023 via email

@xiaofeng94
Copy link
Owner

Hey,

我们总是使用训练结束得到的最后一个模型权重进行测试。训练iterations一般detectron2默认的90k。如果您想选择更优的权重,可以考虑使用验证集进行选择。

@guigui6666
Copy link
Author

guigui6666 commented Aug 8, 2023 via email

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

2 participants