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

Can't reproduce the results of lisa-7b #135

Open
L1uShuai opened this issue Jul 9, 2024 · 4 comments
Open

Can't reproduce the results of lisa-7b #135

L1uShuai opened this issue Jul 9, 2024 · 4 comments

Comments

@L1uShuai
Copy link

L1uShuai commented Jul 9, 2024

Has anyone managed to replicate the outcomes of lisa-7b? I'm puzzled as to why, after ten epochs of training, my test results for giou are 0.05 and for ciou are 0.03.

@L1uShuai
Copy link
Author

L1uShuai commented Jul 9, 2024

I've employed 6 GPUs, and the llava version I'm using is LLaVA-Lightning-7B-delta-v1-1. And the rest of the settings have been left as is.

@wxliii
Copy link

wxliii commented Jul 30, 2024

Have you solved this problem?

@L1uShuai
Copy link
Author

Have you solved this problem?

Up until now, I haven't been able to resolve this issue.

@chenqi1126
Copy link

Maybe you should add the delta (LLaVA-Lightning-7B-delta-v1-1) to the original LLaMA weights to obtain the LLaVA weights (LLaVA-Lightning-7B-v1-1). Refer to the legacy models part in https://github.com/haotian-liu/LLaVA/blob/main/docs/MODEL_ZOO.md

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