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

碰到一个非常头疼的问题,根据odvg的数据集格式自制了一个数据集,一万多张图,配置好了可以进行训练,但是训练过程中一直没有精度,特别低总是0.00几,跑了几十个epoch也是这样,谁有碰到类似的情况吗?球球了,帮帮忙,跪谢跪谢! #65

Open
junnnxie opened this issue Apr 11, 2024 · 10 comments

Comments

@junnnxie
Copy link

No description provided.

@funny000
Copy link

现在还是这个问题吗,会不会是数据内容处理的不对

@jokercao6
Copy link

train的精度我也不会改,但我把val的mAP能展示出来

@HZWHH
Copy link

HZWHH commented Jun 4, 2024

请问问题解决了吗?

@jiajihao
Copy link

batch size 调小点试试

@TaoTXiXi
Copy link

你好,请问你解决了么,我现在也是遇到这个问题

@sipie800
Copy link

是不是amp训练,dino常常会有这个问题

@TaoTXiXi
Copy link

我发现我是在转coco格式的时候类别索引号加了1,所以出现了这种情况,类别索引号应该从0开始

@Nahtreom
Copy link

我也遇到了同样的问题,请问你解决了吗?

@Xuxiusheng
Copy link

您好,请问您解决了吗

@ysl1402
Copy link

ysl1402 commented Sep 25, 2024

您好,请问您解决了吗,我也遇到了同样的问题

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

10 participants