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

在单卡GPU上训练出现的问题,无法解决 #268

Open
lyzla opened this issue Jun 2, 2024 · 3 comments
Open

在单卡GPU上训练出现的问题,无法解决 #268

lyzla opened this issue Jun 2, 2024 · 3 comments

Comments

@lyzla
Copy link

lyzla commented Jun 2, 2024

作者您好,我用的Imagenet-mini数据集,在单卡GPU上训练。看到您之前回答的问题:在单卡GPU上训练,需要将lr减小到原来的8倍,这样设置:--nproc_per_node=1、--lr 2.5e-4。但是我尝试了这样设置的前提下,batchsize=128
image
batchsize=64
image
batchsize=32
image
batchsize=16
image
代码没有任何修改的情况下,都会出现不同程度的问题。

@lyzla lyzla changed the title 你好,请问你用的是单张卡来训练吗?lr是多少?batchsize是多少呀?我也在imagenet-mini上训练,设置的参数是这样的--lr 2e-4、-b 16,但是loss很大 在单卡GPU上训练出现的问题,无法解决 Jun 2, 2024
@iamhankai
Copy link
Member

这是什么模型

@lyzla
Copy link
Author

lyzla commented Jun 3, 2024

@zhangxuliqiang
Copy link

您好,我想问一下ghostnet的imagenet预训练模型是不是在model文件夹里面自带的那个

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