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 accuracy on the dataset version-2 #11

Open
toggle1995 opened this issue Sep 30, 2023 · 3 comments
Open

The accuracy on the dataset version-2 #11

toggle1995 opened this issue Sep 30, 2023 · 3 comments

Comments

@toggle1995
Copy link

Hi, thanks for you amazing work.
I find you release an update dataset with only noncompound images. So can you tell us the accuracy on this new version-2? I think the accuracy reported in the your paper is evaluated on version-1?

@xiaoman-zhang
Copy link
Owner

For the model pre-trained on the version-2 dataset, the accuracy on the multiple-choice task is 0.587, and on the open-ended blanking task is 0.343.

@ShawnHuang497
Copy link

For the model pre-trained on the version-2 dataset, the accuracy on the multiple-choice task is 0.587, and on the open-ended blanking task is 0.343.

Hi, could you tell us the model architecture details about these metrics like Table 3 in your paper?

@AmeeraBawazir
Copy link

For the model pre-trained on the version-2 dataset, the accuracy on the multiple-choice task is 0.587, and on the open-ended blanking task is 0.343.

Hey, this evaluation result is for which model variant, the MedVInt-TE or MedVInt-TD?

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