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

关于参数设置对结果的影响? #7

Open
moontsar opened this issue Sep 7, 2022 · 1 comment
Open

关于参数设置对结果的影响? #7

moontsar opened this issue Sep 7, 2022 · 1 comment

Comments

@moontsar
Copy link

moontsar commented Sep 7, 2022

image

1、表13给出了不同数据集使用了不同的参数设置,如果想在自己的数据集上调优,请问优先调整哪些参数呢?
2、训练半监督之前需要先进行有监督模型的训练,请问为什么半监督用了8卡,而有监督只用了2卡训练,如果有监督用8卡是否预训练模型指标更高呢?还是说LabelMatch方法需要的有监督预训练模型不需要很强时,性能提升比使用更强的预训练模型时,指标更高?
最后感谢你们开源的优秀工作!

@AceCoooool
Copy link
Collaborator

  1. 参数部分我们没有进行具体的设置;
  2. ① 主要是有监督部分的数据量比较少,采用8卡的话需要增大epoch数目;这块具体采用8卡或者2卡其实都是可以的,只是8卡情况下具体epoch我们没有试过具体多少更合适 ② 半监督里面有监督预训练模型有影响,但在半监督训练的时候迭代次数一般比较大,这个影响因素不算大

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