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

关于参数调节 #15

Closed
jianjiandandande opened this issue Oct 5, 2020 · 6 comments
Closed

关于参数调节 #15

jianjiandandande opened this issue Oct 5, 2020 · 6 comments

Comments

@jianjiandandande
Copy link

jianjiandandande commented Oct 5, 2020

可以分享一下单目标跟踪里面关于参数调节的一些技巧吗

@jianjiandandande jianjiandandande changed the title About train epoch 关于参数调节 Oct 5, 2020
@HonglinChu
Copy link
Owner

HonglinChu commented Oct 7, 2020

关于参数调节,商汤开源的pysot里面,有一个hp_search.py的文件用于自动调参;可调节的参数有:惩罚因子,学习率,窗口影响因子,搜索区域大小等

@jianjiandandande
Copy link
Author

好,我试试

@HonglinChu
Copy link
Owner

HonglinChu commented Oct 9, 2020 via email

@jianjiandandande
Copy link
Author

好的,谢谢啦

@jianjiandandande
Copy link
Author

Alexnet version are 20 epochs,   resnet50 version are 50 epochs

------------------ 原始邮件 ------------------ 发件人: "HonglinChu/SiamTrackers" <notifications@github.com>; 发送时间: 2020年10月5日(星期一) 下午4:38 收件人: "HonglinChu/SiamTrackers"<SiamTrackers@noreply.github.com>; 抄送: "Subscribed"<subscribed@noreply.github.com>; 主题: [HonglinChu/SiamTrackers] About train epoch (#15) The number of training times of SiamRPN++ in the original paper is 20 epochs, but you have trained 40 and 50 epochs. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or unsubscribe.

On the contrary, Alexnet version are 50 epochs,   resnet50 version are 20 epochs, I looked at the source code carefully

@HonglinChu
Copy link
Owner

HonglinChu commented Oct 13, 2020 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