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

val.pt和test.pt的区别以及评估时的问题 #85

Open
WanyanHuang opened this issue Mar 20, 2024 · 3 comments
Open

val.pt和test.pt的区别以及评估时的问题 #85

WanyanHuang opened this issue Mar 20, 2024 · 3 comments

Comments

@WanyanHuang
Copy link

cityscapes pidnet-s 对应两个权重,val.pt和test.pt,请问两个权重是如何训练的?猜测val.pt是用train做训练集,val做验证集,test是train+val做训练集,val做测试集。另外,评估时,是不是都是用val进行评估的?目前用val做测试集,val.pt的mIoU为78.76%,test.pt是86.96%(比官方的77.8高了很多),是否正常?

@djl00
Copy link

djl00 commented Mar 31, 2024

你好,我想问一下,你在使用官方给的权重测试时遇到这个问题了吗,我不知道该怎么解决了,期待你的回复:
image

@que-shen
Copy link

请问两个权重的问题你解决了吗

@1404561326521
Copy link

你好,我想问一下,你在使用官方给的权重测试时遇到这个问题了吗,我不知道该怎么解决了,期待你的回复: image

这是因为模型的权重参数包含在“state_dict”字典内,在读取时加上['state_dict']就可以正常读取了,torch.load(cfg.MODEL.PRETRAINED['state_dict'])

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

4 participants