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

The module name of mask decoder is not the same as the name of mask decoder in SAM2. #102

Open
CloudyDory opened this issue Dec 6, 2024 · 2 comments

Comments

@CloudyDory
Copy link

Hi, when running the SAM2-Adapter training code, I notice that the name of mask decoder in SAM2-Adapter is self.mask_decoder, but in the original SAM2 the module is named self.sam_mask_decoder. Therefore, when loading the SAM2 checkpoints in line 191 of train.py, the trained parameters of the mask decoder is not loaded, and is therefore trained from scratch in SAM2. I hope to know if this is the indented behavior of the code?

@tianrun-chen
Copy link
Owner

Kindly want to confirm which model of SAM2 are you using? We are using https://dl.fbaipublicfiles.com/segment_anything_2/092824/sam2.1_hiera_large.pt

@CloudyDory
Copy link
Author

Hi, I am also using this file.

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

2 participants