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.

Sign up for GitHub

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

Jump to bottom

资源占用问题 #32

Open
Ming-Di opened this issue Sep 18, 2023 · 3 comments
Open

资源占用问题 #32

Ming-Di opened this issue Sep 18, 2023 · 3 comments

Comments

@Ming-Di
Copy link

Ming-Di commented Sep 18, 2023

请问在RLHF过程中，actor，refrence，critic和reward使用的都是7B吗，使用offload了吗，我用的4张80G卡，使用offload的情况下，加载完模型就占用60g了，batch size=4，显存就占满了
@Ablustrund
Copy link
Collaborator

您好，我们没有使用offload。加载完模型每个卡占用60G，尝试适配一下Zero3。应该需要修改少量的代码（50行以内）。

@Ablustrund
Copy link
Collaborator

我们在开源代码时去掉了Zero3的部分。经过验证代码是可以完美适配Zero3的。

@Ming-Di
Copy link
Author

Ming-Di commented Sep 21, 2023

您好，我们没有使用offload。加载完模型每个卡占用60G，尝试适配一下Zero3。应该需要修改少量的代码（50行以内）。

非常感谢😁，请问您训练时 是使用8卡 + zero3 + 不开offload 的设定吗，这样的话 batchsize能到多少呢，我想参考一下

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@Ablustrund @Ming-Di