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

value of the output temperature is 0.005078125. In the paper, I see it is 1.3. Why is this? #44

Closed
orrzohar opened this issue Sep 10, 2023 · 1 comment
Assignees

Comments

@orrzohar
Copy link
Owner

          Hi @orrzohar During the reasoning process, when I execute class PostProcess, the value of the output temperature is 0.005078125. In the paper, I see it is 1.3. Why is this?

Originally posted by @YH-2023 in #43 (comment)

@orrzohar orrzohar self-assigned this Sep 10, 2023
@orrzohar
Copy link
Owner Author

Hi,
The reason is that I devide the temperature by the hidden_dim:

postprocessors = {'bbox': PostProcess(invalid_cls_logits, temperature=args.obj_temp/args.hidden_dim)}

I do this so the objectness prediction won't be as dependent on the hidden_dim of the transformer model. As hidden_dim=256,
1.3/256 = 0.005078125

Best,
Orr

@orrzohar orrzohar mentioned this issue Sep 10, 2023
Closed
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

1 participant