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

请教一下batch_size的设定是否对小样本分割有影响 #56

Open
xia-yu66 opened this issue May 24, 2023 · 5 comments
Open

请教一下batch_size的设定是否对小样本分割有影响 #56

xia-yu66 opened this issue May 24, 2023 · 5 comments

Comments

@xia-yu66
Copy link

作者你好,请教一下,我用3090的显卡可以设置成batch_size为16。但是我看其他论文即使是用的性能更好的显卡,但是batch_size普遍设置成4或者8,我想问一下batch_size的大小对小样本分割的结果影响大嘛?感谢!

@chunbolang
Copy link
Owner

有影响但不显著,voc数据集大概在1个点左右。如果改动不是很多的话一般与baseline保持相同设置即可。在方案确定后也可以尝试做一下消融确定适合自己网络的超参。

@xia-yu66
Copy link
Author

xia-yu66 commented Jun 5, 2023

好的 谢谢了

@xia-yu66
Copy link
Author

xia-yu66 commented Jun 7, 2023

抱歉又要打扰你了,我对两阶段的训练方式不太理解,比如在spilt0(1-5类别),第一阶段训练完成之后,是生成了6-20类别的分割图继续用于第二阶段吗,我这样理解的对吗

@chunbolang
Copy link
Owner

是的 训练了能够提供基类预测的base-learner

@LonelyPlanetIoT
Copy link

请问如果我自meta learner 里面做改动 是否还需要重新训练base learner 呢?

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