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

How to Load a Close-set Fine-tuned Grounding Dino into Grounded SAM #498

Closed
narchitect opened this issue May 14, 2024 · 4 comments
Closed

Comments

@narchitect
Copy link

Hello. Following issue #241, I fine-tuned a grounding Dino via mmdetection for a close-set. However, when I try to load these fine-tuned weights into Grounded SAM, it says the model layers are different, and thus, loading fails. I'm curious if there's a way to successfully load a close-set fine-tuned Grounding Dino model into Grounded SAM. It seems it would be highly effective to be able to load the close-set model for downstream tasks. Thanks in advance for your response.

@rcao-hk
Copy link

rcao-hk commented May 28, 2024

Same issue, look at the network defination is slightly different.

@shz0519
Copy link

shz0519 commented Sep 24, 2024

Hello, how to solve this problem?

Same issue, look at the network defination is slightly different.

@narchitect
Copy link
Author

Hey @rcao-hk and @shz0519, thanks for your attention, and sorry for the late reply. I’ve resolved the issue by creating new code to perform transfer learning on both models using MMDetection's Grounding DINO and the SAM model. I referenced the Grounded SAM code and modified the bounding box extraction with MMDetection's DINO model.

@yoshi-show1
Copy link

Hey @rcao-hk and @shz0519, thanks for your attention, and sorry for the late reply. I’ve resolved the issue by creating new code to perform transfer learning on both models using MMDetection's Grounding DINO and the SAM model. I referenced the Grounded SAM code and modified the bounding box extraction with MMDetection's DINO model.

Thank you for the information. If possible, could you share the method? I am facing the same issue and am struggling with it

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