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

Keypoints inputs issue for learned interiors (Numerical Calibrator) #12

Open
brianpark0126 opened this issue Dec 7, 2023 · 0 comments · May be fixed by #13
Open

Keypoints inputs issue for learned interiors (Numerical Calibrator) #12

brianpark0126 opened this issue Dec 7, 2023 · 0 comments · May be fixed by #13
Assignees

Comments

@brianpark0126
Copy link
Collaborator

brianpark0126 commented Dec 7, 2023


Describe the bug:

keypoints_inputs() method does not properly update, and instead runs on the previous iteration of keypoints constructed during forward pass.


To Reproduce
Steps to reproduce the behavior:

  1. Create a numerical calibrator with learned input keypoints
  2. Train the module on a few examples
  3. call calibrator.keypoints_inputs()
  4. Return value will be from the values from before the final training step.

Expected behavior
The final training step updates should be properly referenced. When using learned input keypoints, the keypoints_inputs function should generate the input keypoints directly from the logits.

Additional context
N/A

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

Successfully merging a pull request may close this issue.

1 participant